A Social Identity Perspective on the Management of Confidential Information in Organisational Contexts Kiran Singh BPsych (Hons)

A thesis submitted for the degree of Master of Philosophy at The University of Queensland in 2017 School of Psychology

ii Abstract No organisation or profession can succeed without confidentiality. There can be no health- care, legal representation, research, journalism or even nations without confidentiality and the trust that underpins it. Yet despite the critical role confidentiality plays in all these professions, managing confidentiality is recognised as an increasingly difficult challenge in the modern world. For this reason, gaining an understanding of the psychological processes that feed into the management of confidential information is more vital than ever. To date, it is clear that psychologists have paid little attention to the group processes that are implicated in people’s willingness to preserve, or else violate, confidentiality. Relatedly, it is apparent that there is little investigation of whether, and to what extent the exchange of confidential information is implicated in people’s development of a shared identity. This suggests that the social dimensions of confidentiality demand empirical examination — particularly as these relate to the expression and development of shared social identity. This is the goal of this thesis. First, it seeks to provide a conceptual analysis of how and why issues of social identity are both a product of, and a precursor to, understanding confidentiality in the workplace. Second, it presents two studies that support this conceptual analysis. Study 1 shows social identity impacts individuals’ management of confidential information; Study 2 shows how the act of sharing confidential information can serve to build a sense of shared social identity. A key conclusion here is that while confidentiality discourse has predominantly taken an individual focus that implicates personal skills or traits as the basis for preserving or breaching confidentiality, I argue that there is a need to also consider the ways in which group dynamics (and the associated identity-related processes) dictate the way confidential information is perceived and acted upon. Theoretical and practical implications are discussed, as well as future directions for the field.

iii

Declaration by author This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

iv Publications during candidature No publications.

Conference papers:

1. Singh, R, K., Haslam., A., & Steffens, N. (2015, April). The social psychology of confidentiality. Presentation at the Annual Convention of the Society for Personality and Social Psychology, Newcastle, Australia.

2. Singh, R, K., Haslam., A., & Steffens, N. (2014, November). The role of social identity in directing our decisions about confidential information. Paper presented at the University of Queensland Research Higher Degree Day, Brisbane, Australia.

3. Singh, R, K., Haslam., A., & Steffens, N. (2014, November). Building strong ties: receiving and sharing confidential information enhances shared identity. Paper presented at the University of Queensland Centre of Research in Social Psychology Lab Group, Brisbane, Australia.

Publications included in this thesis No publications included.

v Contributions by others to the thesis My primary advisor, Professor Alex Haslam provided assistance in understanding the theoretical background to the studies as well as how to frame the research narrative and provided comments on thesis drafts. My secondary advisor, Dr. Nik Steffens commented on results from both studies and provided insight into running follow-up mediation analyses. Data analysis strategy and all statistical analyses in both studies were conducted by myself as the researcher.

Statement of parts of the thesis submitted to qualify for the award of another degree None.

vi Acknowledgements

“You have been my friend,” replied Charlotte. “That in itself is a tremendous thing...after all, what’s a life anyway? We’re born, we live a little while, we die...By helping you, perhaps I was trying to lift up my life a trifle. Heaven knows anyone’s life can stand a little of that”.

E.B. White (Charlotte’s Web, 1980, emphasis added])

There is a story behind this thesis. Although it is comprised of my own work, it represents a mosaic of all the people that have come either before or after but are united in leaving their mark. First and foremost, I express my immense gratitude to my supervisor, Alex Haslam. Alex — your resolute passion to research, your unnerving attention to detail and your unwavering dedication in bringing out the best work in your students represent only a handful of your qualities. I thank you, for the time and effort you invested in shaping me as a researcher. My thanks also extend to my other supervisor, Niklas Steffens for his enthusiasm in interpreting data and his commitment in seeing this project to the very end. My sincere thanks must also include my chair, Courtney von Hippel. Courtney — your ability to be ever so patient throughout the process and your super-human powers to instill confidence in your students is nothing short of admirable. Finally, I must thank Alison Pike for all her efforts in facilitating this entire process and being a genuinely warm source of support when one needed a kind ear. To my dear friend Samuel Sparks. Sam — I cannot thank you enough for all that you have done for me. You have been a friend to me in every sense of the word that the opening quotation echoes. I hope we remain friends for even longer than forever. Similarly, from our fitness sessions where we endlessly talked about the rest of our lives, to those moments we shared as we tried to navigate research waters, I learnt one key thing. It is possible to be a true friend and a good researcher. Anne Sheldon, you are both. To those friends who I might not have the opportunity to see as often, but who have, nonetheless been instrumental in providing support along the way — Shelli Dubbs, Fiona Barlow, Cagla Sanri, Tom Norton, Dan Crimston, Jeremy Nash, Zoe Walter, Sean Murphy and of course, Shel Leveque. I thank you sincerely, for your time, energy and for making my research experience a pleasant one. To my significant others outside of the world of psychology — to my friends and family — I cannot express to you how grateful I am. In particular, I must thank Lavanya Sukumar for being there through it all and for never failing in making me laugh at the darkest of times. This is a true testament to your character. A special thanks to my family. To my dad and mum, the important things in life — hope, courage and happiness — were things that you both helped me find again. You represent the best of me. To my father and mother-in-law for their endless encouragement and

vii support. And to my dear siblings, Danny, Simba and Bean, for reminding me of the importance of not taking life too seriously — that there is time to dance and be silly. Always. Finally, in a thesis that speaks to the idea that our social relationships shape our realities, it is nothing short of remiss for me not to acknowledge the one person that has profoundly shaped my own. To my husband, Amandeep Sarai for his support, guidance and his unwavering faith in my abilities and for providing me with the most powerful thing in the whole world — unconditional love.

viii Keywords social identity, self-categorisation, organizations, confidentiality, communication.

Australian and New Zealand Standard Research Classifications (ANZSRC) ANZSRC code: 170113 Social and Community Psychology 80% ANZSRC code: 170107 Industrial and Organizational Psychology 20%

Fields of Research (FoR) Classification FoR code: 1707 Psychology 100%

ix Table of Contents

Abstract ...... ii

Declaration by author ...... iii

Publications during candidature ...... iv

Publications included in this thesis ...... iv

Contributions by others to the thesis ...... v

Statement of parts of the thesis submitted to qualify for the award of another degree ...... v

Acknowledgements ...... vi

Keywords ...... viii

Australian and New Zealand Standard Research Classifications (ANZSRC) ...... viii

Fields of Research (FoR) Classification ...... viii

Table of Contents ...... ix

List of Figures ...... xii

List of Abbreviations used in this thesis ...... xii

Chapter 1 ...... 13

Introduction and overview ...... 13 Setting the Scene: Contextualizing the Present Research ...... 14 Statement of the Problem ...... 15 Aims, Scope and Overall Structure of Thesis ...... 17

Chapter 2 ...... 19

Psychological approaches to confidentiality management ...... 19 Overview ...... 20 Defining Confidentiality ...... 20 The Difference Between Confidentiality and Privacy ...... 20 Defining Confidentiality Management ...... 21 Psychological Approaches to Confidentiality Management ...... 21 Perceptions and management of confidential information...... 21 Who breaches confidentiality? ...... 23 Preventing breaches of confidentiality ...... 24 Summary ...... 24

Chapter 3 ...... 26

x Social identity, communication and confidentiality ...... 26 Overview ...... 27 Introduction ...... 27 The Social Identity Approach ...... 28 Social Identity is a Determinant of Communication ...... 31 Desire to share information...... 31 Responses to receiving information...... 32 Social Identity as a Byproduct of Communication ...... 34 Social identity and shared cognition...... 34 Effective communication and job satisfaction...... 36 Identification builds job satisfaction...... 37 Social Identity and Confidential Information ...... 38

Chapter 4 ...... 43

Study 1: Evidence that shared social identity impacts on the management of confidential information ...... 43 Overview ...... 44 Background ...... 44

Method ...... 46 Participants and Design ...... 46 Procedure and Manipulations ...... 46

Results ...... 49 Simple correlations ...... 51

Discussion ...... 52 Summary ...... 53

Chapter 5 ...... 55

Study 2: Evidence that sharing confidential information creates shared social identity ...... 55 Overview ...... 56 Background ...... 56

Method ...... 58 Pre-testing ...... 58 Participants and Design ...... 59 Procedure and Manipulations ...... 59

Results ...... 61 Data analysis strategy ...... 61

xi Discussion ...... 63 Summary ...... 64

Chapter 6 ...... 66

General discussion and conclusions ...... 66 Introduction and Overview ...... 67 Social Identity as a Precursor to Confidentiality Management ...... 67 Social Identity as a Product of Confidentiality Management ...... 70 Broader Implications ...... 72 Limitations and Future Directions ...... 73 Concluding Comment ...... 74

References ...... 76

xii

List of Figures

Figure 1. Number of registered data leaks from the period of 2006 to 2014. Adapted from Infowatch Analytical Center, Global Data Leakage Report (2014). Report of the Infowatch Analytical Center on the Global Data Leaks in 2014...... 15 Figure 2. The role of shared social identity (salient self-categories) in determining how individuals think about, what they do with, and how they interpret confidential information...... 41 Figure 3. Perceived information utility as a function of social identity and designated confidentiality...... 50 Figure 4. Study 1: Intragroup motivations to pass on information as a function of social identity and designated confidentiality...... 51 Figure 5. Study 2: Perceived confidentiality as a function of designated confidentiality ...... 62 Figure 6. Social identification mediates the relationship between designated confidentiality and job satisfaction...... 63

List of Abbreviations used in this thesis

SCT– Self-categorization theory SIT – Social identity theory ANOVA – Analysis of variance

Chapter 1 Introduction and overview

Chapter 1 14 “Last month, I spent the better part of five days at a Florida hospital sitting at the bed-side of my very old father. In the bed next to him, there was a man who had been hospitalised after passing out from unregulated diabetes. His daughter couldn't bring him home with her, and she was attempting to have him transferred to a Veterans' Administration hospital, since he did not have private health insurance. In the room across the hall was an elderly gentleman. He had an inoperable brain aneurysm. He was unconscious, and the doctors said he needed to be put on life support. His wife said he didn’t want that type of heroics, but his son was insisting that everything possible be done. Down the hall, a woman in her 50s was told that her breast cancer had spread to other organs. The doctor was recommending more surgery and a major course of radiation and chemotherapy ... Everything I’ve told you I heard from hallway conversations between medical personnel as I sat bedside with my dad”. Charles Inlander (Marketplace, 1999 [emphasis added])

Setting the Scene: Contextualizing the Present Research In the process of writing this thesis, I stumbled across the above transcript by Charles Inlander — a health commentator on Public Radio International's Marketplace and president of the largest consumer health advocacy organisation in the United States. Much to my surprise, these words echoed observations I had made long ago during my own work in a clinical setting. My role as a recreational officer involved working with a variety of allied health professionals to bridge the gap between patients and staff. What I found, though, was not what you might expect. You might think that the information patients’ disclose to their doctor is handled with discretion and that confidentiality is preserved. This, however, did not seem to be the case. In fact, there were many instances where patient information was shared among staff in ways that not only violated confidentiality but also seriously undermined privacy. These decisions, however, did not appear to occur in isolation. Within the hospital environment, the exchange of confidential information takes place in a highly stratified context. Increasingly, the provision of health care is no longer contingent upon a unique doctor–patient relationship but rather, most treatment models involve cross-disciplinary collaboration. Specifically, confidential information that concerns the patient is handled by a group of health practitioners that differ not only in their profession, occupation, status and rank but also in the professional ethical guidelines that surround the disclosure of information. In short, confidential patient information is managed in the context of individuals embedded in a complex social system in which many of the parties belong to different groups. Given all this, it seemed pertinent to ask whether, and how, these social factors influence confidential information exchange not just in hospitals, but generally in organisations that manage sensitive information. Indeed, it was these initial suspicions that formed the impetus for a thesis

Chapter 1 15 dedicated to investigating the social dimensions of confidentiality in the organisational domain. Statement of the Problem The protection of information is a fundamental challenge in today's digital world. In the Western world at least, the idea that individuals often disclose information that is intended to be protected is not new. Anecdotal evidence attests to this fact. Mainstream media is ridden with these types of scandals — hackers that access online personal details, whistle-blowers that lift the lid on the dealings of their employers, and, at the apex of these cases, individuals who leak information related to significant issues of national security. The scale of this growth is evidenced by the findings of a recent annual report by InfoWatch Analytical Centre. This report aims to provide a comprehensive picture of confidential information data leaks across different countries including, but not limited to, United States, Australia, Great Britain and Russia. Findings suggest that since 2006, the number of registered information leaks have been on the rise, with the latest statistic showing a 22% rise in 2014 compared to the previous year (see Figure 1). In short, confidential information leaks proliferate throughout the world. It seems likely too that these trends will only increase in the years ahead (e.g., see Pouillot, 2001).

1600 1395 1400

1200 1143

1000 934 794 801 800 747

600 530 Number of leaksof Number

400 333 198 200

0 2006 2007 2008 2009 2010 2011 2012 2013 2014

Figure 1. Number of registered data leaks from the period of 2006 to 2014. Adapted from Infowatch Analytical Center, Global Data Leakage Report (2014). Report of the Infowatch Analytical Center on the Global Data Leaks in 2014.

The observation that confidential information does not tend to remain confidential represents a very real paradox that exists throughout Western society. Indeed, scholars from an array of

Chapter 1 16 disciplines, including philosophers, have observed, “our culture, is a culture of disclosure, but it is also a culture of confidentiality… to disclose and to keep secret, to expose and to hide" (Cordess, 2001 [emphasis added]). This paradox rears its head, in no small way, in professional spheres where individuals are expected to handle sensitive information. It is little wonder, then that these professionals often find themselves conflicted about whether or not they should breach the confidentiality of information they have obtained from their client or patient in the provision of services (e.g., Adler, 1995; Baird & Rupert, 1987; Cordess, 2000; Cummings, 2000; DeBernardo, & Shoemaker, 1998; Kämpf, & McSherry, 2006; McDonald, 1990; Pope & Vetter, 1992; Steinfield, Ekorenrud, Gillett, Quirk & Eytan, 2006; Trowell, 2001). It makes sense that scholars describe the discourse of confidentiality as involving far more than a simple blanket prohibition on sharing information. In sum, confidentiality implicates two countervailing forces. On the one hand, professionals are subject to increasingly stringent ethical guidelines that seek to protect the confidentiality of their client (e.g., to safeguard their interests, and preserve their autonomy). At the same time, legal guidelines dictate an increased flow of confidential information (i.e., potentially for the welfare of the individual and third parties; Clark, 2006; Kämpf, McSherry, Thomas, & Abrahams, 2008; Kirkpatrick, Lopez & Prosser, 1999; Vogel, & Nyman, 2011). All of these details highlight the fact that issues of confidentiality are increasingly salient and complex. Despite all of the streams of evidence that speak to the importance of understanding confidentiality, this topic has only recently attracted interest and gained momentum in the psychological literature. Indeed, from the early 1990's to the present day, the number of research articles whose titles, abstracts of keywords reference "confidentiality" or "confidential communications" has increased by 800%. In the empirical research that has been conducted to date, psychologists have generally focused on examining (a) the effects of assuring confidentiality, (b) the processes that lead to breaches of confidentiality, and (c) the consequences of such breaches (e.g., for social relationships and organizational performance; Brann & Mattson; Clarke, Cott, & Drinka, 2007; Cline & McKenzie, 2000; Derlega, Metts, Petronio, & Margulis, 1993; Dorodny, 1998; Dubois, 2002; Hook & Cleveland, 1999; Paterson, & Mulligan, 2003; Shardlow, 1995; Singer, Dawn & Esther, 1995 for a review; Whetten- Goldstein, Nguyen & Sugarman, 2001). Despite strong growth in such research, relatively few studies have been conducted within the field of social psychology (at least in comparison to other sub-disciplines, notably organisational and managerial psychology). Typically, too, psychologists study confidentiality management at the individual level, and the role of group memberships (and the social identities underpinning them) remain unexplored. Intuitively, this individual-level approach makes sense. Ultimately, the decision to disclose confidential information is made by an individual. I will argue, however, that this

Chapter 1 17 approach is too narrow because it fails to capture the social group-based underpinnings of confidentiality management. Moreover, progress in this area is hindered by the absence of a guiding theoretical framework, and so existing approaches do little to advance our understanding of why people make particular decisions about the management of confidential information in social and organizational contexts. Such insights would have considerable utility beyond the field of psychology. Accordingly, this thesis seeks to meet a threefold purpose: first, to make theoretical progress by examining the social dimensions of confidentiality; second, to provide empirical evidence of the importance of these dimensions; third, to use this evidence to outline future paths for research in the field that will be of both theoretical and practical importance. Aims, Scope and Overall Structure of Thesis The aim of this thesis is to provide theoretical and empirical insights into how a social identity approach feeds into confidentiality management, and to test whether this approach provides a more comprehensive picture of the psychological processes that are involved in individuals' decisions surrounding confidential information at work. More particularly, the thesis aims to extend upon prior literature by arguing that how people manage confidentiality is always a characteristic of a specific social relationship and therefore depends on people’s motivations towards one another, and on the way these are shaped by their sense of shared (or non-shared) group membership (Haslam, Reicher, & Levine, 2012). If we are to understand how individuals manage confidential information, I suggest that it is critical to recognize that people working together in the same group or culture may make decisions about confidential information on the basis of how they relate to one another — in much the same way that a multidisciplinary team handles confidential information that concerns the patient. Thus, the management of confidential information cannot be understood exclusively at the individual level, but also needs to attend to group-level realities. Spelling this point out further, more research is required to determine whether group processes are implicated in people’s willingness to preserve, or else violate, confidentiality. Additionally, it will be important to determine how the exchange of confidential information has implications for the individuals and the group — and in particular, the sense of social identity that people derive from being part of a psychological group (Tajfel, 1972). Indeed, I will argue that without attending to the role that groups and group dynamics play in shaping confidentiality management, our understanding of when and how people come to their decisions with confidential information will be deficient in key respects. In these terms, the key theoretical premise underpinning this analysis is that our understanding of the process of confidentiality management needs to appreciate, and take account of, intergroup dynamics (e.g., after Sherif, 1967). Moreover, given the centrality of confidential communications to all organisations, the

Chapter 1 18 practical implications of rectifying such an omission are considerable. Most particularly, organisations would gain important insight into how to direct employees who are managing confidential information so as to achieve optimal outcomes for all parties (Dyer & Noeoka, 2000; Hale & Boyrs, 2013; Harwood; Turnock & Ashleigh, 2014; James, Leiblein & Lu, 2013; March, Schulz & Zhou, 2000; Robinson, 1991; Stiles & Petrila, 2011). In terms of thesis structure, theoretical chapters will first review existing literature that points to the different ways in which scholars have approached issues of confidentiality. This literature reviewed in Chapter 2 provides some important insights, but, it is clear that it also has distinct limitations — most obviously in failing to address the ways in which the management of confidential information is structured by unfolding social relationships. Accordingly, in Chapter 3 I provide the case for a theoretical analysis which focuses on the ways in which social identity processes are implicated in confidentiality management. To this end, I review relevant literature that speaks to the core proposition of this thesis: that social identity is both a precursor and a product of confidentiality management. More particularly, I argue that social identity (i.e., an internalized sense of shared group membership; Tajfel, 1972; Turner, 1982) plays a key role in individual’s construal of confidential information and also that the sharing of confidential information can create a sense of shared social identity. Chapters 4 and 5 will provide systematic empirical examination of this theoretical proposition. More specifically, Study 1 explores the role of social identity in individuals’ management of confidential information and Study 2 explores how the act of sharing confidential information can build a shared social identity. Finally, Chapter 6 will integrate the findings of these studies and provide a general discussion that details both the implications and limitations of this work and identifies important avenues for future inquiry.

Chapter 2 Psychological approaches to confidentiality management

Chapter 2 20 “What I may see or hear in the course of the treatment or even outside of the treatment in regard to the life of men, which should not be noised abroad, as to this I will keep to myself, holding such things unfitting to be spoken”. from the Hippocratic Oath, Hippocrates, c 400 b.c.

Overview Confidentiality is hardly a new concept. Philosophers have long recognised that physicians bear an ethical and professional duty to safeguard their patients’ information. Indeed, arguably the most renowned reprimand against a breach of confidentiality appears in the philosophical writings of Hippocrates over two millennia ago — that have become the cornerstone of medical ethics today (as epitomized by the Hippocratic oath; Markel, 2004). The purpose of this chapter is to review psychological perspectives and how they can help us understand issues of confidentiality management. First, I will first outline key definitional issues identified in the literature, then clarify how the concept is distinct from privacy and finally, provide a conceptual definition of confidentiality management. Thereafter, I will discuss psychological accounts of confidentiality management with an emphasis on perceptions and actions. The chapter then concludes by outlining contributions and limitations of this literature. Ultimately, the goal here is to set the scene for the introduction of a new framework that seeks to understand (and empirically examine) confidentiality management from a social psychological perspective. Defining Confidentiality Before continuing, it is useful to define the key concepts with which this thesis is concerned. The Latin term for trust, confidere, lies at the root of two closely related notions of 'confidentiality' and 'confide' (Larkin, Moskop, Sanders & Derse, 1994). Although many scholars agree that trust is central to defining the term, there is limited consensus on a conceptual definition and some commentators argue that this has hindered empirical progress on the topic (e.g., Beauchamp & Childress, 2000; Clark, 2006; Gillion, 1984). In this work, note that confidentiality is defined in a more general sense. Collocating the literature also reveals another conundrum: the interchangeable use of the term 'privacy' with 'confidentiality'. In what follows, it is these matters which are addressed first. The Difference Between Confidentiality and Privacy In their seminal review article, Warren and Brandeis (1890) define privacy simply as the "right to be let alone" and it thus refers to individuals' ability to restrict or control their personal information prior to sharing that information (for more extensive reviews on the historical background of privacy see the review by Pritts, 2008; see also Westin, 1967). Confidentiality is narrower in scope than privacy. More particularly, it is a mode of managing the fate of shared

Chapter 2 21 information that ensures that the individual has control over who is permitted access to this information (Beauchamp & Childress, 2001). Breaches of confidentiality refer to unauthorised or accidental disclosure of information to third parties, without obtaining explicit permission (Gostin & Hodge, 2002). Thus, confidentiality seeks to specify the terms of a communication ("Information about me being protected and not widely shared") whilst privacy broadly connotes a particular status ("I don't want anyone to know that"; Clark, 2006; Prewitt, 2001). Defining Confidentiality Management Broadly speaking, in professional contexts, confidentiality is conceptualised as an ethical principle that governs the disclosure of sensitive information (Allen, 1997; David & Kitchener, 2008). In particular, this means that information will not be disclosed to third parties without informed consent or in ways inconsistent with the original disclosure except under conditions agreed to by the source (Bersoff, 2008; Cain, 1998; Larkin, et al., 1994; see also Australian Institute of Health and Welfare, 2003). Importantly too, Allen (2008) alerts us to the means through which confidentiality is made possible, that is through discretion, silence, and ensuring that data is secure. And so, in summary, confidential communications are those in which a significant matter has been communicated, the matter necessitates professional silence, and the label connotes trust that the recipient will act on it appropriately (Clark, 2006). The term 'management' in this thesis refers to both perceptions (e.g., under what conditions people perceive information as 'confidential') in addition to behaviour (e.g., individuals' tendency to share information that is identified as confidential). Psychological Approaches to Confidentiality Management Perceptions and management of confidential information. This section of the thesis will explore the perspectives of professionals and clients with regard to how confidentiality is conceptualised and how it is treated within the professional world. Within the psychological tradition, the majority of the research focuses on perceptions of confidentiality from the patient or client’s perspective and, as such, is studied primarily in medical and legal fields. In understanding how individuals construe confidential information, a subset of researchers has, in one way or another, simply asked participants what the label of confidentiality connotes. For example, a number of studies show that — at least in the medical domain — people understand confidentiality to convey a general expectation that 'something that was done or said' is restricted from general disclosure (Brann & Mattson, 2004; Jenkins, Merz & Sankar, 2005; see also, Lyren, Kodish, Lazebnik, O'Riordan, 2005; Magnusson, Oakley & Townsend, 2007; McKinstry, Watson, Pinnock, Heaney & Sheik, 2009). Similarly, other evidence within the legal context demonstrates that individuals perceive the label of confidential to mean that information is 'just between you and me' or 'it's private, and no

Chapter 2 22 one should know' (Lyren, Kodish, Lazebnik & O’Riordan, 2005; Michelle, Paulsel, Virginia, Richmond, McCroskey & Caymans, 2005; Sankar, Moran, Merz, Jones, 2003, for a review; Whetten-Foldstein, Nguyen & Sugarcane, 2001). One of the most obvious points derived from this literature is that most people generally do have some basic or straightforward understanding of what confidentiality means and entails. Nevertheless, a key point to be made here is that there is considerable variability in how people perceive confidential information should be treated by professionals. To this end, individuals’ expectation that their matters are held in absolute confidence does not necessarily hold true in reality. For example, survey data collected by Ormrod and Ambrose (1999) revealed that — when clients were inquired as to whether their discussions with their professionals are subjected to confidentiality — an overwhelming majority perceived a greater degree of confidentiality than what was actually the case in practice. Building on this recognition that there is a great deal of discrepancy between the matters that clients perceive are to be kept confidential and what is actually treated as confidential, other research also points to the fact that this discrepancy is seen in the content itself. For example, studies show that some patients tend to broadly conceptualise the majority of information they convey to their doctor to be 'confidential' whereas a subset of patients that perceive the term to be applicable only to censored topics (e.g., mental illness or sexual activity; Carlisle, Shickle, Cork & McDonagh, 2006; Cheng, Savageau & Sattler, 1993; Gosling, Potts, Handelsman, 2000, Isaacs & Stone, 2001; Jenkins, 2005; Jenkins, Merz, & Sankar, 2005). Along related lines, empirical work in the clinical field demonstrates that the majority of patient populations think that confidentiality only applies to the assessments and treatments aspect of health care (and not just 'everything' they tell their doctor; Magnusson, et al., 2007; Samet, Winter, Grant & Hingson, 1997). Thus, these different streams of evidence point to the idea that although people understand confidentiality to relate to the protection of information, it is less clear which (or whether) specific information is implicated (for a detailed discussion on this, see Clarke, 2006). In answering how people handle confidential information, the literature has primarily focused on professional practitioners' perspectives. Here it is apparent that despite a level of consistency regarding guidance about confidentiality in practice, handling confidential issues is far from straightforward — a point that is confirmed by a considerable body of both basic and applied research across numerous sub-disciplines (Davis & Mickelson, 1994; Ford, Millstein, Halpern- Fisher & Irwin 1997; Isaacs & Stone, 1999; Ledyard, 1998; Pérez-Cárceles, Pereñiguez, Osuna & Luna, 2005). For example, studies indicate that the most serious ethical issue to emerge in professional practice is the decision to break confidentiality (Jacob-Timm, 1999; Pope & Vetter,

Chapter 2 23 1992; Younggren & Harris, 2008) and relatedly, that issues of confidentiality are often brought up for systematic review (Stapleton, Douglas, Hays, & Parham, 2010). In attempting to understand why confidential information is difficult to manage, researchers have found that there is little consensus amongst professionals on the key issues: (1) What exactly represents harm, (2) who should be protected, and (3) the conditions under which it is ethically appropriate to disclose information to third parties (Fennig, Barak, Benyakar, Farina, Blum & Treves, 2000; Ormrod & Ambrose, 1999 & Pinta, 2009). For example, in a study by Duncan, Williams and Knowles (2012), in 43% of cases, Australian psychologists were unable to reach consensus as to whether a breach of confidentiality was warranted or appropriate (see also, Isaacs & Stone, 1999; 2001; Pendrak & Ericson, 1998 and Taylor & Adelman, 1998). Similarly, a study conducted by Davis and Michelson (1994) found that school counsellors reached agreement in just over half of the ethical dilemmas presented to them and that consistency in breaching confidentiality only emerged about suicide. In short, deciding the best course of action is fraught with difficulties (see Davidson & Davidson, 1996). Thus, as Sullivan and colleagues (2002) explain, "there is a tension or push–pull relationship between the factors, such that the Negative Nature of the Behavior factor may push one toward a decision to break confidentiality, whereas the Maintaining the Therapeutic Process factor may pull one away from breaking confidentiality" (p.399). A review of the literature on perceptions of confidentiality, and on the motivations underpinning these perceptions, thus supports Bersoff’s conclusion that '[e]xcept for the ultimate precept – above all, do no harm – there is probably no ethical value in psychology that is more inculcated than confidentiality.… There is probably no ethical duty more misunderstood' (2003, p.155). Who breaches confidentiality? It is perhaps not surprising that researchers have tailored their investigation to specify the conditions responsible for the differing outcomes of breaching versus protecting confidentiality. Specifically, research has focused on a variety of personal characteristics of confidants and informers including demographics (i.e., age, gender, the level of ethics education) and other characteristics (i.e., organisational tenure) as well as contextual factors. One of the more intriguing findings in this regard is that familiarity with the ethics of confidentiality is associated with greater propensity to breach it (Baird & Rupert, 1987; Elger, 2009). Relatedly, in medical and educational contexts, this propensity is heightened to the extent the client is young, female and involved in risk-taking behaviour (e.g., alcohol consumption) that increases in intensity (e.g., consuming relatively little versus a lot) and/or frequency (e.g., annually to several times in a month; Rae, Sullivan, Peña Razo, George, & Ramirez, 2002; see also Moyer & Sullivan, 2008). Moreover, the whistle-blowing literature demonstrates that — compared to inactive

Chapter 2 24 witnesses — whistleblowers tend to have high levels of education and job performance and score higher on tests of moral reasoning (see Magnus & Viswesvaran, 2005 for a detailed meta-analysis). Preventing breaches of confidentiality. In order to circumvent the potential ramifications of breaches of confidentiality in the health-care context, researchers have advocated a range of different approaches. In particular, Fisher (2008) proposes a sequence of six steps in an ethical model that illustrates the significance of the principle, specifies the rules and exceptions, and encourages psychologists to reach for the “ethical ceiling” (p.4). For example, the first step is to accentuate the importance of confidentiality to the client and the sixth is to stress confidentiality to peers and students. Similarly, Van Lieu (2012) sets out a series of more arguably practical recommendations, suggesting that professionals (a) inform clients which staff will have access to their records and (b) restrict access to information to certain health care providers. While such strictures are intuitively appealing, the narrow practical perspective that informs them neglects the psychological processes that contribute to confidentiality management. Along related lines, on a practical level, researchers have commented on the unrealistic nature of the idealised picture of confidentiality presented in guidelines on the provision of health care services. As Hughes and Louw (2002, p.14) observe:

“Guidelines do not adequately address the reality of confidentiality in modern practice... perhaps because its legalistic, professional ethics have a philosophically limited view of confidentiality…. a more philosophical picture regards people as embedded in a shared, worldly context in which relationships and mutual engagement become crucial”.

In making this observation, Hughes and Louw might thus be characterized as advocates for an approach to confidentiality that is inherently social. Summary Although psychological approaches to confidentiality provide some important insights into issues of confidentiality, they are somewhat limited in their scope. In particular, I argue that these psychological accounts do little to enhance our understanding because they focus on the rules, ethics, and legalities that surround confidentiality. Although these rules and guidelines are part of the social structure that influence decisions about confidentiality, I contend, that such are not necessarily the whole picture. To complement this analysis, I argue that research should be oriented to the processes by which, and through which, confidentiality-related decisions are made. Fundamental to this is the recognition that the management of confidential is grounded in, and depends on upon, our social relationships. In short, confidentiality is a function of relationships

Chapter 2 25 (Cordess, 2001). And so, I argue that the issue of confidentiality management needs to be understood (and addressed) through the lens — and methods — of social psychology.

Chapter 3 Social identity, communication and confidentiality

Chapter 3 Overview In this second theoretical chapter, I will provide a brief introduction to contextualise the literature that will be reviewed. After this, I will introduce core principles of the social identity approach and discuss a number of relevant theoretical contributions that have been informed by this approach (Social Identity Theory [SIT]: Tajfel & Turner, 1979 and Turner, 1985) (Self- Categorization Theory [SCT]: Turner, 1985; Turner, Hogg, Oakes, Reicher & Wetherell, 1987). Thereafter, I will outline how this approach has invigorated fresh research within the management literature. These streams of literature examine the bidirectional effects of social identity and effective communication within the workplace. More specifically, this section of the chapter explores (1) the key role social identity plays in the unfolding dynamics of organisational communication and (2) the identity-relevant consequences of the communication process (along lines suggested by Morton, Wright, Peters, Reynolds & Haslam, 2012). Importantly too, these streams of literature are reviewed with the overall goal of establishing whether the insights gained from social identity research are also useful in developing and testing hypotheses about confidentiality management. Having done so, the final section of this chapter presents a framework for meaningfully extending upon this literature by developing a model which specifies how the effects of group life and confidentiality management interact with one another. Ultimately too, we will explore whether and how this model has the capacity to enhance our understanding of confidentiality management within workplace contexts. Introduction “No matter how secure a system is, there’s always a way to break through. Often, the human elements of the system are the easiest to manipulate and deceive”. Hadnagy (2011, p. xv) The opening quotation from Christopher Hadnagy draws our attention to the idea that social psychology may matter as much as physical structures in the emergence of confidentiality breaches. Although this idea is not new, in recent years, a new body of research has emerged that specifically examines the way in which social and group forces shape and contour patterns of communication (e.g., Bourhis, 1991; Gallois, Ogay & Giles, 2004; Giles, 2008; Hardwood & Giles, 2005; Postmes, Tanis & Wit, 2001; Suzuki, 1998). A separate (although arguably, related) body of research has provided extensive contributions on the other side of the coin: exploring the way in which communication itself impacts the formation and maintenance of our social relationships (e.g., Koudernberg, Postmes & Godijn, 2011; Swaab, Postmes & Spears, 2008; Postmes, Haslam and Swaab, 2005). Accordingly, this chapter will synthesise the literature within this field. In doing so, it attempts to provide a

Chapter 3 28 conceptual analysis of how and why issues of social identity are both a precursor to, and a product of, understanding confidentiality in the workplace. Prior to understanding how this is the case, though, it is important first to outline key tenets of the social identity approach. The Social Identity Approach The social identity approach incorporates two linked theories: social identity theory (SIT; Tajfel & Turner, 1979) and self-categorization theory (SCT; Turner, 1982; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987; see Haslam, 2004; Postmes & Branscombe, 2010, for recent discussion). Historically, social identity theory was developed to account for a series of findings in early work on intergroup processes of discrimination and conflict. More specifically, Tajfel and colleagues demonstrated that mere categorisation of people into arbitrary groups (e.g., on the basis of their preference for a particular abstract artist) could be sufficient to provoke intergroup antagonism (Tajfel, Flament, Billig, & Bundy, 1971). In a set of experiments, participants were randomly assigned to one of these two groups and their task was to allocate points to either a member of their own group (someone that liked the same artist; i.e., an 'ingroup member') or a member of another group (someone that liked a different artist; i.e., an 'outgroup member'). Intriguingly, participants tended not to distribute these points equally between the two groups, but rather typically chose to allocate more points to ingroup members than to outgroup members. Social identity theorists argued that this tendency to think and behave in ways positive towards a salient ingroup (i.e., exhibit high levels of ingroup favoritism) enabled participants to achieve a positive social identity. Indeed, this speaks to one of the key assumptions of the social identity approach: that in a range of social contexts in which our sense of identity derives from our group membership people seek to positively differentiate their ingroup (“us”) from relevant outgroups (“them”) (Tajfel & Turner, 1979). On the basis of these 'minimal group studies' Tajfel reasoned that we can understand how trivial group allocation impacts on individual level behaviour once we understand how individuals define themselves within any a range of social contexts. In this way, this approach argues for a model of self that is multifaceted — in that it can be defined in terms of one’s individuality and/or in terms of shared group memberships (Amiot, Sablonniere, Terry & Smith, 2007; Haslam, 2004). At an individual level of abstraction, I can define myself as 'I' or 'me' and see myself in terms of my idiosyncratic characteristics. However, at a social level of abstraction, I can define myself as 'we' or 'us' in terms of the characteristics that I share with others by virtue of our common group membership (Haslam, 2004). Self-definition — either as an individual in terms of personal identity or as a member of a group in terms of social identity — has important implications for how individuals perceive, respond, and interact with their social world (Turner, 1982). In particular, when people define

Chapter 3 29 themselves in terms of a meaningful connection to their ingroup, perceived shared group norms and values serve to shape individual-level decisions regarding how to feel or think and also determine what is the appropriate course of action in a particular context (Reicher, Spears & Haslam, 2010; Terry & Hogg, 1996). For example, a soldier who defines himself as a soldier will look to other comrades in order to determine what is expected of him and see feedback from that ingroup as relevant in deciding how to respond to issues that impacts them as a group (e.g., how to respond to particular stressors; Haslam, O’Brien, Jetten, Vomedal & Penna, 2005). Following up on Tajfel’s initial ideas, Turner and colleagues developed and further extended on these ideas in self-categorisation theory (SCT; Turner, et al., 1987, 1994). These scholars sought to understand how people come to share group-based identity, as well as the consequences that this has for their behaviour (van Dick & Haslam, 2012). In this regard, self-categorisation theory asserts that the shift from personal to social identity is underpinned by a process of depersonalization (Turner, 1982) Critically, depersonalization does not connote a loss of self but rather, that our sense of self becomes more inclusive and perceived to be interchangeable with the broader ingroup category (Turner, 1982). Notably, self-categorisation of this form is distinct from the objective groups we belong to on the basis of particular qualities in which individuals may be members of, but not necessarily identify with (e.g., a female engineer in a male-dominated work-place may be more inclined to identify with her profession, rather than her gender). Relatedly, another key contribution of SCT is to identify and explain the determinants of this process of depersonalization— that is, clarifying when we identify with particular group. To this end, self-categorisation theorists argue that social identity salience is a product of principles of fit and perceiver readiness (Oakes, Haslam, & Turner, 1994). These mean that people are more likely to define themselves as a member of a specific group to the extent that they see themselves to 'fit' a particular group and are psychologically 'ready' to be part of that group. Thus, a lawyer is more likely to self-categorise as a lawyer to the extent that (a) the self-categorisation has historical relevance for her (e.g., because she has been a lawyer for a long time), (b) she finds herself in a situation (e.g., court of law) whereby intragroup differences are perceived to be smaller than intergroup differences (the principle of meta-contrast; Turner, 1985; such that she sees other lawyers to be similar to herself), and (c) the group membership is meaningful in some way (because the lawyers are prosecuting a case). However, it is also true that people belong to a range of different groups and that this is likely to have complex implications for cognition and behaviour. In addressing this proposition, SCT asserts that we behave differently as a function of the specific identity that is made salient within any given context (e.g., Haslam, Morton, Haslam, Varnes, Graham & Gamaz, 2012). Thus, a lawyer in a court of law might exhibit behavior consistent with the norms of lawyers (such as being

Chapter 3 30 disciplined, objective and firm) but her behaviour as a sports enthusiast at a cricket ground could be very different (perhaps being unruly, partisan, and relaxed). In thinking about the importance of context, SCT also argues that there are times when we think of our social identity as more exclusively defined at the subordinate level (Turner, 1982). However, there are also situations in which our self-concept becomes more inclusively defined at the higher superordinate level. For example, studies show that a supporter of a rival football team can be construed as an outgroup member (e.g., as a Liverpool or Manchester United supporter) in one context but in another context may be categorized as an ingroup member at a higher level of abstraction (e.g., as a football supporter; e.g., Gaertner, Mann, Murrell, & Dovidio, 1989; Levine, Prosser, Evans & Reicher, 2005). Together, these processes of flexible self-categorization allow for context-dependent definitions of self that can allow individuals to understand and orient to others in variable ways and thereby function adaptively within a changing social world (Haslam, 2001). Since its emergence roughly two and half decades ago, the social identity approach has provided a platform for theoretical and empirical progress in a range of disciplines. The research this has led to is diverse but united in its focus on the way in which different aspects of group life (and the identities that underpin them) play a role directly, or interact with other factors, in determining behaviour. For example, there has been ongoing interest in exploring the ways in which social identity (and factors associated with it; e.g., solidarity, social connection, a sense of community) make an independent contribution to individuals’ health and wellbeing (termed ‘the social cure’; see Jetten et al., 2012 for further details). Most relevant to our present purposes, though, the social identity approach has provided a novel framework for understanding various forms of organisational behaviour. This includes, but is not limited to, work on motivation, group decision-making, productivity, stress, leadership, and cooperation (e.g., Duck & Fielding, 1999; Ellemers, Gilder & Haslam, 2001; Haslam & Platow, 2001; for overviews see Haslam, 2001; van Knippenberg, 2000). The relevance of a social identity approach derives from the fact organisations are highly stratified entities in which individuals belonging to different groups, with whose members they must co-ordinate their efforts, in the context of working with one another to achieve work-related goals (Scott, 1997; Trice & Beyer, 1993). Because organisations are inherently group contexts (Ashforth and Mael, 1989), various scholars have also outlined the implications this has for how people communicate at work (e.g., Gardner, Paulsen, Gallois, Callan, Monaghan, 2001; Paulsen, Graham, Jones, Callan & Gallois, 2005; Haslam, 2001). This literature is a natural starting point from which to start our investigation. This is because breakdowns of confidentiality can be conceptualised fundamentally as failures of communication arising from a particular form of

Chapter 3 31 organizational disconnection — in particular, reflecting a lack of shared understanding (e.g., regarding norms and expectations; Paulsel, Richmond, McCroskey & Cayanus, 2005). Social Identity is a Determinant of Communication Desire to share information. To understand why and how confidentiality management is structured by shared group membership, it is instructive to consider how the social identity approach informs our understanding of communication dynamics more generally. As a starting point, consider the fact that group structures are central to all organizations. Employees are differentiated from one another on the basis of their group memberships— for example in terms of their workgroup, team, profession, functional role, division, department, and employment status. This highly differentiated structure means that employees can — and frequently do — define their sense of self in terms of social identity (as 'we ' and 'us') (over and above their personal identity as 'I' or 'me'). This social self-definition is then fundamental to how they relate to their work with others (see Gardner, et al., 2001). In particular, this is because identity-based relationships (in which identity is shared or non-shared with others) matter for how people communicate, what is communicated, and how it is subsequently understood (Morton et al., 2012). In particular, because they see themselves as sharing the same perspective on the world and the same interests (Turner, 1991), individuals will generally be more motivated to share information with people whom they perceive as members of the same salient social category (i.e., ingroup rather than outgroup members). Importantly too, the information that is exchanged (i.e., its content) is likely to be relevant to aspects of their shared group membership. For example, in the context of health-care, the behavior of two members of a hospital might be structured by their common profession (e.g., as doctors) and so not only will they be motivated to talk to one another, but also the content of their conversation is likely to be relevant to that identity (e.g., how to treat a particular patient). The idea that people are generally more motivated to share information with members of their psychologically salient group ('us') rather than the outgroup comparison ('them') has been empirically supported in both experimental and in survey studies. For example, studies have found that employees tend to communicate more openly with members of their own work-team than with members of other teams (e.g., Agama, 1997). Other evidence shows that people communicate differently as a function of external categories that are not necessarily relevant to the organization but that may become salient in a particular context— for example, on the basis of common national identity within multicultural workplaces (e.g., Suzuki, 1998). Conversely, when people do not perceive themselves to share identity with others (e.g., where there is conflict between long-standing employees and new trainees; Levine & Moreland, 1999), the motivation to exchange information tends to be diminished (e.g., McGarty, Haslam, Hutchinson, & Turner, 1994). Indeed, studies also show that those who neglect this psychological

Chapter 3 32 connection do so at their peril because non-shared social identity is often the basis for miscommunication and misunderstanding (e.g., Petronio, Ellemers, Giles, & Gallois, 1998; Weick & Roberts, 1993). Thus, as Haslam (2001) argues that, on their own, formal requirements for people talk to one another (e.g., as part of a performance appraisal) may often be ineffective— because here people talk to each other because they must, not necessarily because they want to. The most basic idea to take away from these ideas is that generally; people will be more inclined to share information with members of their group rather than members of the other group. However, an important caveat to this idea is that the processes of communication also change meaningfully as a function of the nature of the current intergroup relationships as well as the content of the ingroup’s social identity (Livingstone & Haslam, 2008). Although there are times that we might be generally more willing to share information with others who are perceived to share group membership with us, this does not always equate to a disinclination to share information with outgroup members. More particularly, the tendency to pass on information to ingroup members rather than outgroup members may be accentuated in the case of a contentious intergroup relationship but attenuated in the context of a cooperative one (e.g., Agama, 1997). Along similar lines, the content of an ingroup’s social identity is likely to be manifested in specific norms, and as such, upholding these norms might require providing information to outgroup members. As such, if an ingroup has norms of aiding and looking after patients, then conforming to such norms should translate into motivations to communicate with relevant outgroups, when it serves these goals (e.g., such that doctors are happy to pass on medical information to nurses). Responses to receiving information. In the preceding section, the clear focus was on how identity-related processes are implicated in whether or not people are motivated to communicate with another person, as well as how they communicate. What previous research shows very clearly is that these two aspects of communication vary as a function of shared (or non-shared) identity structures. Our analysis now shifts to the insights the social identity approach brings in understanding the way in which individuals respond to communication. Here early work by Mackie and colleagues (1990) demonstrated that the extent to which a message was accepted by a recipient depended upon the communicator’s group membership. In a series of experiments, students at the University of California, Santa Barbara (UCSB) were informed that messages pertinent to university testing either emanated from an ingroup member (a fellow student at UCSB) or an outgroup member (University of New Hampshire, UNH). In addition to this social identity manipulation, participants were presented with messages that either conveyed strong or weak points in relation to the subject matter. Unsurprisingly, students tended to be more persuaded by the strong argument compared to the weak one. Importantly, though, this

Chapter 3 33 effect was conditioned by the group membership of the person providing the message such that if the message was provided by an outgroup member, the message had minimal influence regardless of the strength of the argument— ostensibly because it could no longer be presumed to be accurate or worth taking seriously (for related findings see McGarty, Haslam, Hutchinson & Turner, 1994; Wilder, 1979). More recent work has also focused on using social identity theory as a platform from which to understand the psychological impact of message content. For example, studies show that an individual’s willingness to accept criticism (and to enact change based on it) differs as a function of whether the critic is an ingroup member (‘an insider’) or an outgroup member (‘an outsider’). This idea— referred to as the intergroup sensitivity effect — is empirically supported by a studies that demonstrate individuals construe the input of others positively, only to the extent that comments ostensibly emerge from fellow members (Hornsey & Imani, 2004; Hornsey, Oppes & Svensson, 2002). The researchers argue that the intergroup sensitivity effect emerges on the basis of a perception of different motives. Whereas outgroup critics tend not to be tolerated and to be met with defensiveness and suspicion, ingroup critics appear to receive “the benefit of the doubt” and are presumed to be acting in the interests of the group (for a more detailed reviews of outgroup criticism, see Morton, et al, 2012; Rabinovich & Morton, 2010). As a result, individuals are more likely to enact change in response to criticism from fellow ingroup members. Amongst other things, this is because shared group membership provides a shared cognitive framework from which to perceive and act upon information (Postmes, 2003). Consequently, ingroup members— those with whom one shares social identity— are perceived to be trustworthy and far more influential than outgroup members— those that one does not share social identity with (Foddy, Platow, & Yamagishi, 2009; Greenaway, Wright, Willigham, Reynolds, & Haslam, 2015). One implication of this is that information received from an ingroup member is likely to be perceived in the manner that it was intended but also be acted upon, in meaningful ways that may facilitate organisational outcomes (Ford, Henderson, & O’Hare, 2013; Weick & Roberts, 1993). Collectively then, these ideas suggest that group membership may not only structure who we are motivated to share information with, but which information is considered worthy of sharing. Thus, individuals might be particularly motivated to share information (generally) with ingroup members to the extent that it comes from an ingroup member rather than an outgroup member. In summary, social identity influences and shapes individual-level motivations to communicate, to pay attention when receiving information, and to respond to communication in meaningful ways. In other words, categorization of the self into important groups and identification with those groups proves to be an important determinant of communication dynamics (Haslam,

Chapter 3 34 Jetten, O'Brien, & Jacobs, 2004). Thus, the key contribution of this approach to understanding communication dynamics is that group membership, and the sense of identity associated with that membership (as 'we' and 'us') acts as a pivotal driver of the fate and flow of information throughout the organization (Morton et al., 2012). Social Identity as a Byproduct of Communication The foregoing discussion focused on how social identification plays a role in communication dynamics. In this section, the focus shifts to how communication dynamics shape and contour patterns of social identification. Although these two parts of the chapter are the opposite sides of the same coin, they are united in their main objective— namely, to lay the groundwork for a conceptual analysis of how and why these identity-based relationships are crucial for understanding and resolving issues of confidentiality management. In the popular 1960’s comedy, Get Smart, an unskilled spy in a meeting with his boss insists on using an entirely defective product to discuss matters regarded as classified information. During these discussions, the “cone of silence”— a transparent device in the shape of two conjoined fishbowls descends from the ceiling to enclose the heads of the spy and the boss. The device’s cumbersome construction provides for entertainment, but it also serves as an analogy. Specifically, it speaks to the idea that some extra form of closeness results when people disclose particular types of information (such that, they have their own communication “bubble”). In what follows, we review research which asks more formally whether and how the exchange of confidential information can craft a sense of ‘togetherness’ in which the recipient feels that they share social identity with the source of the information. Although this idea is yet to be examined systematically from a social identity perspective, research on shared cognition provides a range of important pointers. Social identity and shared cognition. Up to this point, our analysis of social identity theorising has predominantly focused on the way in which salient social identities impact on cognition and behaviour. However, these ideas are also relevant to the question of how individuals within a group context can also interact with one another to form a shared cognition that can fuel a sense of shared identity. In particular, research in the shared cognition literature presents evidence which suggests communication is an important way in which individuals develop common group membership (Swaab, Postmes, van Beest, & Spears, 2007; Swaab, Postmes & Spears, 2008). When individuals attempt a task collectively in small groups, prior to the task, levels of uncertainty are often high. For this reason, individuals should be particularly predisposed to converse with one another as a means of reducing uncertainty (Turner, 1991). This opportunity to communicate can help them to reconcile their differences and express informal group norms (Cannon-Bowers, & Salas, 2001). Moreover, this process of conversing with other individuals

Chapter 3 35 provides the foundation for the emergence of a shared cognition (Klimoski & Mohammed, 1994). Shared cognition refers to "sharedness and congruence of knowledge structures that may exist at different levels of conceptualization within a group"(Swaab, Postmes, Beest, Spears, 2007). This means that when individuals talk to one another they are able to form a shared understanding and not only “be on the same page” but also to a get a sense of “us” and “who we are” (Haslam, et al., 2012, p. 162; see also Postmes, et al., 2005). For example, a new book club might develop a set of standards which dictate the books they decide to read, and the process of reading and talking about these should in turn feed into the content and strength of the book club’s identity. One implication of this analysis is that communication is more than just a process that involves a give-and-take of words and meaning. This is evidenced by the fact that lapses in conversation — even momentary ones — can be accompanied by an unsettling feeling and a need to quickly “fill in the space” because they appear to disrupt the flow of interaction (Koudernberg, et al., 2011). In this way, we see that the communication process itself has psychological effects that feed into processes of identity formation and expression. Speaking directly to this point, research in the communication literature has found that conversation that is natural and fluent is linked to a range of positive emotions. This positive emotion heightens an individual’s perceptions of belonging to the group, which in turn, may give rise to processes of social identification (e.g., Rabiowitz, 2008; Rosenthal & Knee 1994; Koudernberg, et al., 2011). Thus, effective communication orients individuals towards psychologically fusing with one another because it helps to satisfy basic human needs (e.g., for self- esteem and social validation; see Gaertner, Iuzzini, Guerror, Witt & Orina, 2006; Pratt, 1998). We can hypothesise too that sharing confidential information may be particularly powerful in this respect. Giving and receiving valuable, sensitive information seems likely to increase feelings of trust and togetherness and thereby lead to a heightened sense of group identification. Other empirical findings are also consistent with this notion that sharing confidential information, in particular, might induce heightened feelings of social identification. For example, although workplace gossip is not necessarily equivalent to confidential information, gossiping with colleagues is one of the contexts in which confidential information can be shared (Baker & Jones, 1996; Nicholson, 2001). Relevant research in this tradition has broadly examined how gossip — an “evaluative exchange of information about an absent third party” — serves a number of a key social functions (e.g., McAndrew & Milenkovic, 2002; Wilson, Wilczynski, Wells, & Weiser, 2000). In particular, it is apparent that the organisational “grapevine” in which gossip is communicated plays a valuable role (a) in the expression of tacit group norms, (b) in heightening self-esteem, (c) in gauging trustworthiness of others and (d) in creating bonds within social groups (Baumeister, Zhang & Vohs, 2004; McAndrew, Bell, & Garcia, 2007; Wert & Salovey, 2004). Because gossip

Chapter 3 36 functions to help us become socialised into our professional role and adapt accordingly to the work ethos, it comes as no surprise that people make an active effort to talk to others on their first day “on the job” (i.e., for more on the process of socialisation for newcomers, see Saks & Ashforth, 1997; see also Laing, 1993). Particularly pertinent to confidentiality, those employees who are privy to the “office gossip” may perceive that they are part of the workgroup and experience a sense of mutual trust and respect (e.g., as “dependable juniors”). Conversely, those who do not, might perceive that they are excluded by the group and simply there to fulfill obligatory work requirements (e.g., “hired hands”). Taken together, these empirical findings speak to the capacity of gossip to form distinct insider/outsider group dynamics. The idea that sharing secrets binds individuals together interpersonally has been widely supported in previous research (see Caughlin & Petronio, 2004 for a comprehensive overview). This research leads us to speculate that confidential information (which is akin to gossip in many ways) may also be a group bonding agent. More particularly, we can hypothesise that receiving confidential information heightens a sense of belonging, social validation, and self-esteem which in turn, helps us to form connections with our group and to cultivate a stronger sense of “us” (see also Kurland & Pelled, 2000; Soeters & van Iterson, 2002). Conversely, those who do not receive confidential information from their ingroup might feel that they have been left out and/or that they cannot be trusted. Taken together, these various ideas provide some initial insights into the ways in which the exchange of confidential information might influence identification processes. At this point, it is useful to also elaborate upon these propositions to consider how they may also shape organizational outcomes more generally. Effective communication and job satisfaction. To understand how good communication feeds into higher levels of job satisfaction, it is instructive to review literature within the field of management that examines leader-member exchange. Briefly, from this perspective, the dyadic relationship between a leader and a member of their organisational unit is fundamental to outcomes at the individual and organisational level (e.g., Graen, 1976; Graen & Cashman, 1975). This means, for example, that the way in which leaders relate to, and interact with, their employees is an important factor in predicting job satisfaction and performance. In understanding how this is the case, consider the idea that supervisors have a limited number of resources and as such, may not distribute their resources equally among their employees (Graen & Uhl-Bien, 1995). This means then that some leader-follower dyads will be characterised by high-quality exchanges (e.g., mutual liking, trust, and respect) whereas these aspects will be neglected (or simply do not exist) in low-quality exchanges (Loi, Chan & Lam, 2014; Loi, Mao & Ngo, 2009). One implication of this state of affairs is that employees in the former situation are

Chapter 3 37 likely to feel happier at work than those in the latter. Indeed, meta-analyses have found that it is high quality leader-member exchanges in particular that predict not only individuals' commitment to the organisation but also their job performance and satisfaction (Amos & Herrick, 2005; for a meta-analysis see Gerstner & Day, 1997). Identification builds job satisfaction. Although the above evidence suggests that effective communication is a key predictor of job satisfaction, the mechanism through which this operates is yet to be fully articulated. To this end, in what follows I provide a conceptual justification for the proposition that a sense of social identity-based psychological attachment underpins the positive relationship between effective communication and job satisfaction. In the organisational domain, identification with a particular unit is driven by the desire to satisfy two fundamental human needs: uncertainty reduction and self-enhancement (Pratt, 1998). More specifically, uncertainty reduction speaks to idea that individuals need to belong to a social group, so that they are able to anchor their self-concept and able to understand “who I am”. In learning about their self-concept in group-based terms, individuals benefit because they become certain of how they ought to behave and thus are able to function with assuredness within this context (Hogg, 2000; Turner, 1991). This means that employees who identify with an organisational ingroup and feel that their need to reduce uncertainty is met, are likely to experience higher levels of job satisfaction than those who do not. Organisational identification also contributes to self-enhancement. Amongst other things, this is evidenced by the fact that, if they can, individuals generally try to join organisations (and organisational units) that are seen to be prestigious (Haslam, Powell & Turner, 2000). A key reason for this is that if they are successful in this endeavor, then this is likely to fulfill the need to achieve a positive and distinct sense of identity (Tajfel & Turner, 1979). As such, individuals are likely to report higher levels of job satisfaction to the extent that they feel identified and relatedly, feel proud to be part of a particular group membership. Indeed, this explains why individuals may be happy to work for a high-status organisation even if it is associated with less financial remuneration than working for a low-status one (Haslam et al., 2000). Following on from this sense of pride in group membership, it seems plausible that highly identified individuals will be emotionally invested in their ingroup. This sense of emotional investment should dispose individuals to make more positive primary appraisals at work (Boroş, Curşeu & Miclea, 2011). As a corollary, individuals should be more likely to adhere to organisational goals and to be motivated to coordinate their efforts with others to achieve those goals (Brewer & Kramer, 1986). Again, this occurs because individuals are also more inclined to trust and like their colleagues which, arguably, also makes their overall work experience far pleasant (van Dick & Haslam, 2012). In this way, even in difficult work conditions (e.g., those

Chapter 3 38 associated with bomb disposal) employees can be content with their work life because it is perceived to be part of “getting the job done” and is underpinned by a sense of “we are in this together” (Haslam et al., 2005). Thus far, our analysis has focused primarily on the processes through which patterns of shared identity might develop and on the way in which communication is pivotal to that process. Moreover, we have also seen that this work speaks to the idea that a sense of shared identity might develop via particular forms of communication and that this in turn will lead to higher perceptions of job satisfaction. Yet while this cross-disciplinary evidence provides us with some clues as to the kind of effects we envisage, more systematic examination is clearly required to explore and test the way in which identity is bound up with, and predicated upon, a specific type of communication exchange — namely that of confidential information. Social Identity and Confidential Information As we saw in Chapter 2, researchers who study confidentiality in the workplace tend to focus on interpersonal aspects of the process and, in seeking to foster effective management of confidential information, propose various sets of rules that individuals should follow in order to achieve this. The consensus here is that as long as individuals follow these rules then this, in turn, should lead to improved outcomes for the organisations. However, one problem with this line of thought is that it neglects the fact that individuals do not always think, feel and behave as individuals (particularly, within modern highly differentiated work systems). Instead, as we have seen, they are likely to be part of a salient psychological group, and this collective aspect of their psychology is likely to be important in shaping their thoughts and motivations when it comes to handling confidential information. More specifically, SCT leads us to predict that in intragroup contexts, personal identity concerns will be particularly important and will serve as the cognitive and motivational basis for confidentiality management. However, in intergroup contexts these concerns will be more likely to be shaped by a person’s social identity. This leads us to speculate that in a range of organisational contexts, group dynamics are likely to be an important determinant of confidentiality management dynamics. Thus, although it is not disputed that confidentiality management is personal, it is plausible that it also has social determinants that remain largely unexplored in research to date. As we have seen, a social identity approach is particularly useful in helping us to theorise about the nature and impact of these social determinants. Accordingly, if we think about organisations as intergroup contexts, and reflect on the fact that this is the precisely the context in which sensitive information is typically handled, then attention to this theoretical perspective would seem to afford an opportunity to enrich our understanding of how individuals manage confidential

Chapter 3 39 information at work. In this regard it is helpful to formalise a few key propositions of the social identity approach, before drawing these together schematically in an integrated model. As we have seen, social identity theory predicts (a) that individuals generally strive to achieve a positive sense of self (of ‘who they are’) and (b) that their sense of self will often be defined by the sense of identity that is furnished by group memberships (as “we” and “us”; Tajfel & Turner, 1979). Under these conditions an individual’s sense of self-esteem is linked to the fate of the group (and to that of fellow ingroup members; Reicher, et al., 2010) because those others are seen to be as part of the collective self (Turner et al., 1987). Accordingly, when a particular identity is salient, individuals are expected to act in ways that advance that group’s interests — attempting to make it better than, and different from, salient outgroups on valued dimensions (Ellemers, Spears & Doosje, 2002). All of this leads to the prediction that when a person’s social identity is salient in the workplace the desire to advance ingroup interests will lead them to try to manage confidential information in ways that advantage that ingroup. Spelling this point out further, we hypothesise that ingroup members may be less inclined, or at the very least, reluctant to disclose confidential information that reflects negatively on their group1. Here, they might be particularly motivated to protect their identity against external threats and try to contain the information as much as possible within their ingroup boundary. On the other hand, a very different motive might underpin their intentions if negatively valenced information concerns a relevant outgroup. In this case, individuals might be motivated to disclose this confidential information a means of undermining the outgroup and/or asserting ingroup power2. Nevertheless, an important caveat here is that these patterns may depend on how “us” is defined— as rival groups within the organization, or one organization that includes both groups against another organization (e.g., see Haslam, Eggins & Reynolds, 2003). More generally, SCT leads us to predict that depending on the level of abstraction at which social identity is defined, individuals might be motivated either to preserve the confidentiality of material pertaining to a select few (in the case of a narrowly defined ingroup identity) or else of a wider group (in the case of a more inclusive identity defined at a higher level of abstraction; Turner et al., 1987). As such, individuals are likely to be less inclined to share confidential information of their group with another, depending on how they define themselves within any given work context.

1 It is important to specify that source and identity-relevance of information are not synonymous here (i.e., it is possible that an outgroup member may disclose confidential information concerning the ingroup, or an ingroup member may disclose confidential information concerning the outgroup).

2 Of course, among other things, this will also depend on the broader set of relations of intergroup relations, and on the particular social identity made salient (Reicher et al., 2012).

Chapter 3 40 When identity is defined more narrowly (e.g., doctors), then individuals might be only less inclined to share confidential information to the extent that it emerges from their group member (e.g., another doctor rather than a nurse). However, if identity is defined more broadly (e.g., in terms of allied health staff) then individuals might also be less inclined to share confidential information (e.g., if a nurse provides it)3. All of this points to the idea that self-categorisation in the workplace provides a motivational basis for the handling confidential information in particular ways. However, perceptions of confidentiality are also critical to this process. To this end, another key proposition of self-categorisation theory is that shared social identity furnishes parties to the communicative exchange with a shared cognitive framework that provides a basis for mutual understanding (Postmes, 2003). When people feel a sense of psychological attachment to their group, not only are they cognitively inclined to reach agreement with members of their group but also behaviourally motivated to pursue efforts that conform to the norms of that group (Turner, 1999). One implication of this is that when sensitive information relates to an ingroup, the perceived confidentiality of confidential information is likely to be heightened. This is because the nature of the information means that individuals are here more likely to see the interests of the ingroup as being threatened (i.e., because a valued identity is at risk). As such, not only are they likely to potentially understand confidential information to be confidential, but to treat it as such and to retaliate against unauthorised information disclosure. Thus, this set of ideas leads us to predict that individuals might not also perceive information to be confidential if it emerges from an ingroup member rather than an outgroup member, but also are behaviorally motivated to preserve the confidentiality of that material. In this way, we see that how employees think about confidentiality and what they do with confidential information is likely to be structured by their social relationships in the workplace and, more particularly by patterns of shared and non-shared social identity (Haslam, et al., 2012). When these social relationships are framed by a sense of shared social identity (so that both the receiver and the source are ingroup members), people should be more inclined to see potentially confidential information as ‘confidential’ and relatedly, be more likely to preserve confidentiality (as indicated as HI in Figure 2).

3 Of course, confidentiality itself can also be defined by intergroup boundaries such that a person in a given group may disclose information with the implicit understanding that it not be disseminated further (e.g., as seen when they say “this information is not to leave this room”).

Chapter 3 41

Information source and Information source and recipient do recipient share salient social not share salient social identity identity Perceive confidentiality Less likely to perceive confidentiality Preserve confidentiality Less likely to preserve confidentiality Optimal interpretation of confidential information Sub-optimal interpretation of incoming confidential information

Information Information Recipient Recipient source source

Salient self-category Salient self-categories Boundary for optimal trust, empathy, Boundaries for optimal trust, communication, accommodation, respect accommodation, communication, respect Figure 2. The role of shared social identity (salient self-categories) in determining how individuals think about, what they do with, and how they interpret confidential information.

Chapter 3 On the other hand, when these social relationships are defined by an absence of shared identity (so that both the receiver and the source are outgroup members), then individuals will not only will be less inclined to perceive information as ‘confidential’ but should also be less inclined to preserve confidentiality (as indicated as H2 in Figure 2). To the extent that this is true, I would argue that an understanding of identity-based relationships is fundamental to optimal confidentiality management in which individuals perceive and handle confidential information in a way that is mutually beneficial. The association between social identity and confidentiality management becomes more clear when we consider the processes that underpin the former are also central to the latter. To this end, a further plausible hypothesis is that a sense of shared social identity creates processes of trust, respect, accommodation and communication which are central to the management of confidential information (Foddy, Platow, & Yamagishi, 2009). Spelling this point out further, confidentiality should be a powerful social signal such that when people give another person confidential information, the act of communicating it indicates that the information provider trusts the recipient to treat the information with a certain level of sensitivity and discretion. Indeed, without some level of identity-based trust, it seems that confidential information is unlikely to be shared (unless with the specific intention to mislead). In short, sharing confidential information is a sign of deep trust and as such should help establish intragroup bonds and socially bind individuals together in a sense of ‘we-ness’. As such, the recipient of confidential information that stems from an ingroup member might be particularly motivated to increase their sense of shared identity with the ingroup. Putting all these ideas together, we hypothesise that when people experience this sense of psychological closeness to other members of their group (and when they perceive that the group membership is valued by their organization) they are more likely to trust and respect each other and work hard to manage confidential information effectively collectively. Accordingly, the central hypothesis of this thesis is that confidentiality management (which includes the extent to which one perceives information as confidential, whether one decides to communicate this information and how this, in turn, is interpreted) is structured by identity concerns that are framed by salient group memberships in the workplace. This, then, is the primary hypotheses that the studies reported in the chapters that follow seek to test.

43

Chapter 4 Study 1: Evidence that shared social identity impacts on the management of confidential information

Chapter 4 44 Overview This first empirical study will provide an initial test of the hypothesis that social identity determines the ways in which people manage confidential information in organizational settings. Specifically, it examines whether group identification (as defined by membership of a fictitious work team) and the designated confidentiality of stimulus information influences individuals’ construal of confidential information and their decisions about how to handle it. The study will also explore other potentially relevant variables with the intention of providing a more comprehensive picture of the processes that underpin the hypothesised effects. In what follows, I will briefly reprise the theoretical logic that informs the study (as outlined in the foregoing chapters) before going on to present details of its procedure and findings. Background Research on the way in which people handle confidential information is of significant interest to the public, professional and scientific world. It is, therefore important to understand what factors are involved in individuals' decisions surrounding confidentiality. Although confidentiality management has a lengthy history within the philosophical tradition (as mentioned in Chapter 2), it has only recently emerged as a focus for psychological scholarship. Here, a large body of recent theoretical and research has paid attention to the ways in which individuals’ personal characteristics explain variation in the management of confidential information (e.g., investigating the way in which the informer's level of ethics education predicts whether they breach confidentiality). Notwithstanding the considerable advances made by this research, in the chapters above it was argued that individual-level accounts are limited in scope because they fail to explain the social underpinnings of confidentiality management. In this regard, a social identity approach to confidentiality management would appear to be particularly useful. This is because it can help us understand when, why, and how the social relationships that arise in the context of groups, impact on the way that individuals perceive confidential information and critically, their intentions regarding how to manage confidential information. The aim of this chapter is to provide an initial test of the ways in which issues related to group membership and social identity are implicated in the management of confidential information. Within this chapter, the current study derived a number of key hypotheses from the social identity approach reviewed in Chapter 3. Each of these hypotheses are outlined below, as well as the rationale that underpins them. For the sake of brevity, the rationale is summarised on the basis of the extensive review covered in the previous chapter.

Chapter 4 45 H1: That participants would perceive information to be more confidential if the information source was an ingroup member rather than an outgroup member. This hypothesis derives from a key proposition of self-categorisation theory, namely that shared social identity furnishes parties to the communicative exchange with a shared cognitive framework that provides a basis for mutual understanding (Postmes, 2003). In other words, when the information recipient shares social identity with the source, they are more likely to want to protect the information that the person gives them than is the case when the source is an outgroup member. In part too, this inclination to see potentially confidential information as ‘confidential’ will reflect the fact that common group membership motivates people to trust, communicate and respect one another (e.g., Tyler & Blader, 2000).

H2a. That participants would be less inclined to share confidential information with another person (ingroup, outgroup, or outside the organisation) when the source was an ingroup member rather than an outgroup member4 There are a number of reasons that people might generally be particularly motivated to preserve the confidentiality of material provided by an ingroup member. Most particularly, when people psychologically categorize themselves as a group member, and act in terms of a particular social identity, preserving the confidentiality of ingroup material reflects their desire to protect (and advance) the interests of their ingroup— something that might be compromised by sharing that information with others. Sharing confidential information might also be seen to violate the terms of one’s ingroup membership, and communicate a lack of trust and respect (e.g., Morton, et al, 2012; Rabinovich & Morton, 2010).

H2b. That participants would be more inclined to share general ('non-confidential') information with ingroup members when the source was an ingroup member rather than an outgroup member. This hypothesis reflects the fact that categorization of the self into significant groups and identification with those groups is argued to be an important determinant of communication dynamics (Haslam, Jetten, O’Brien, & Jacobs, 2004). Speaking to the first part of this hypothesis directly, the inclination to generally share information with ingroup members (rather than outgroup members) will, in part, reflect the fact that shared social identity serves as a basis for an alignment in perspective (in terms of one’s holistic worldview) and interests (Turner, 1991). Indeed, this is consistent with a number of studies which show that people communicate more openly with members of their group than with outgroup members (e.g., Agama, 1997; Suzuki). Relatedly, it is

4 The assumption within this hypothesis is that the information concerns the ingroup or a fellow member of the ingroup.

Chapter 4 46 clear that a lack of shared social identity paves the way for misunderstanding and communication failures (e.g., Levine & Moreland, 1999; Petronio, et al., 1998; Weick & Roberts, 1993). The above hypotheses all relate to the idea that confidential information that ostensibly stems from an ingroup member is likely to have a very different psychological impact than information which stems from an outgroup member. As such, the focus for this study is to establish how recipients respond to information about confidential material provided by another individual and observe whether these reactions differ as a function of group membership of the information provider and the perceiver. With this goal in mind, the current study was designed to examine two factors that might impact on perceptions of, and intentions with, confidential information. In particular, participants were presented with information that was described (a) as either confidential or general information and (b) as emerging from a member of the ingroup or outgroup. More specifically, the study employed a minimal group paradigm in which participants were randomly assigned to one of four cells of a 2 (group membership of the information source: in-group vs. out-group) × 2 (information confidentiality: confidential vs general) experimental design. Method Participants and Design Two hundred undergraduate students at a large metropolitan university in Australia participated in the study. At inception, 230 students were approached to take part in the study and 203 agreed to do so (88% response rate). Three participants who failed the manipulation check were excluded from analyses. In line with the guidelines set out by Simmons, Nelson and Simonsohn (2011) a minimum of 20 participants per cell was a focal objective of data collection. Of the total sample size, N=188 (94%) were female, and 12 (6%) were male. Their average age was 24 (SD = 5.24) ranging from 19 to 53 years. The majority of participants identified themselves as Australian (N=117, 59%) and indicated their ethnicity was Caucasian (N=120, 60%). Informed consent was obtained from all participants before participation in the study. Participants received a voucher as a form of compensation and were debriefed upon completion of the study. Ethical clearance was obtained from the Ethical Review Committee within the School of Psychology at the University of Queensland: 14-PSYCH- MOP-13-JS. As noted above, the study employed a minimal group paradigm in which participants were randomly assigned to one of four cells of a 2 (group membership of the information source: ingroup vs. outgroup) × 2 (information confidentiality: confidential vs. general) experimental design. Procedure and Manipulations Participants were recruited primarily through convenience sampling but also invited to

Chapter 4 47 participate via email and social media. Participants were asked to complete the study in a laboratory in order to ensure consistency across contextual factors. Individuals who agreed to participate in the study were asked to read and reflect on a description of a fictitious organisation and then to complete a self-report questionnaire. Specifically, they were provided with the following information:

Global Training Solutions (GTS) is a worldwide training organisation known for its high-quality training programs. GTS has a number of departments— for example, some employees work in the Business Legacy Usability and Engineering ('BLUE') team and other employees work in the Research and Education Development ('RED') team. We ask that you imagine yourself to be a member of this organisation. Specifically, you are part of the Research and Education Development ('RED') team in GTS.

Importantly, all participants were assigned to be members of the 'RED' work department. Subsequently, they read about an interaction and were asked to imagine that this interaction occurred between themselves and another employee at GTS. In addition, they were informed that they would be asked questions about the content of this interaction. It was stressed that they should read the interaction carefully and direct their attention to the content. Participants were also informed that there were no right or wrong answers. A minimal group paradigm was employed to manipulate group membership (Tajfel, et al., 1971). This involved providing participants with the following information:

As part of a recent professional development strategy, GTS has been interviewing prospective employees to hire from outside the company. Jordan works in [the same work team as you (that is, the RED team) [ingroup condition]/a different work team to you (that is, the BLUE team)] [outgroup condition].

The confidentiality of the material given by the other employee was manipulated by providing participants with the following information: The week before the decision is officially announced, Jordan says to you: "I have some information [general information condition]/This is confidential information [confidential information condition]— I know who is tipped to be the successful applicant for the company’s new hire". Jordan proceeds to tell you who it is.

Chapter 4 48

After reading the vignette, participants completed a series of measures designed to explore their perceptions of, and intentions, regarding the information provided by their colleague. Manipulation checks. Group membership was assessed using one item ("Without referring back, please indicate which work team you belonged to"). This item served as a manipulation check. Participants were offered multiple choices in response and asked to select only one (1 = red, 2 = blue, 3 = don’t remember). Perceived confidentiality was assessed using one item (“I believe the information the [colleague] gave to me was confidential”) in light of evidence that shorter scales are not necessarily problematic (Gardner, Cummings, Dunham & Pierce, 1998; Loo & Kelts, 1998; Postmes, Haslam & Jans, 2012; Scott, Morgan, Plotnikoff & Cubans, 2015). Respondents were asked to indicate to what extent they agreed with this statement (1 = strongly disagree to 7, strongly agree). Dependent variables. Perceived utility of the information provided by the source was assessed using a three-item measure ("I believe the information the [colleague] has given me is useful, I believe the information the [colleague] has given me is important" and "I believe the information [the colleague] has given me is valuable"). Respondents indicated their agreement with these statements (on scales from 1 = strongly disagree to 7, strongly agree). A measure of internal consistency revealed intercorrelations between the items to be high, α = .81, suggesting it was appropriate to aggregate the items to create a single scale. Intention to pass on information was assessed with single-item measures for three separate recipients: ingroup, outgroup, and an entity outside of the current organization5. Here the same item, appropriately reworded, measured motivations to communicate with (a) the ingroup ("I would want to share the information [the colleague] gave to me with members of my work team"), (b) the outgroup ("I would want to share the information [the colleague] gave to me with members of the other work team") and (c) an entity external to the organization ("I would want to share the information [the colleague] gave to me with employees of another company"). Respondents were asked to indicate to what extent they agreed with these statements on scales from 1 (strongly disagree) to 7 (strongly agree). Demographics. Participants reported their age, gender, nationality, ethnicity and whether or not English was their first language. Gender was coded as 0 = male and 1 = female. English as a first language was coded as 0 = yes and 1 = no. The study took approximately 20 minutes to complete. After completing the questionnaire, participants were thanked for taking part in the study and debriefed.

5 These three ratings were not combined to form a single scale.

Chapter 4 49 Results Participants' responses were analysed by means of 2 (information source: ingroup vs. outgroup) × 2 (information type: confidential or general information) analyses of variance. Manipulation check. The manipulation check showed that majority (98%) participants correctly identified the work department they belonged to. Participants who failed this check for the key independent variable were excluded from subsequent analyses. Intriguingly though, responses to the confidentiality manipulation check revealed a significant main effect of information type, such that information was generally perceived to be more confidential when the source labeled it as ‘general information’ (M = 5.26, SD = 1.63) than when they labeled it as 'confidential' (M = 4.76, SD = 1.62), F(1,196) = 4.805, p = .030. Dependent variables. Perceived confidentiality. Consistent with H1, there was a significant main effect of information source, F(1,196) = 5.197, p = .024, such that information was generally perceived to be more confidential when the source of that information was an ingroup member (M = 5.27, SD = 1.52) rather than an outgroup member (M = 4.75, SD = 1.71).There was no interaction between information type and information source for perceived confidentiality, F(1,196) = .031, p = .861. Perceived information utility. A significant main effect emerged for information source, F(1,196) = 9.45, p = .002, such that information was generally seen to be more useful when it emanated from an ingroup member (M = 4.10, SD = 1.19) rather than an outgroup member (M = 3.59, SD = 1.14). There was a marginally significant main effect for information type, F(1,196) = 0.39, p = .053. In addition, there was also a significant interaction between source and information type, F(1,196) = 4.08, p = .045. As can be seen in Figure 3, this arose from the fact that in the case of non-confidential information, information was seen as more useful when it ostensibly emerged from an ingroup member (M = 4.22 SD = 1.22) rather than an outgroup member (M = 3.38, SD = 1.27), t(196) = 3.61, p < .001. However, the perceived utility of confidential information did not vary as a function of information source, t(196) = 0.74, p = .458).

Chapter 4 50

Confidential Non-confidential *** 6

5 4.22

3.99 3.81 4 3.38

3 Perceived informationutility Perceived 2

1

0 Ingroup Outgroup

Group membership of information source

Figure 3. Perceived information utility as a function of social identity and designated confidentiality.

Intention to share information6. There were no main effects for intentions to share information with ingroup members, outgroup members, or people outside the organization. However, there was a significant interaction between information type and information source, F(1,196) = 14.23, p < .001. Contrary to H2a, simple effects analysis indicated that intentions to share confidential information with another (i.e., ingroup, outgroup, or outside the organisation) did not differ as a function of whether the source was an ingroup member (M = 3.44, SD = 1.71) or an outgroup member (M = 4.00, SD = 1.79), t(196) = 1.64, p = .102. However, in line with H2b, as can be seen in Figure 4, participants reported a greater intention to share non-confidential information when its source was an ingroup member (M = 4.00, SD = 1.65) rather than an outgroup member (M = 2.74, SD = 1.65), t(196) = 3.69, p < .001. Importantly too, when information came from an outgroup source, participants reported a greater intention to share confidential information (M = 4.00, SD = 1.79) than general (i.e., 'non-confidential') information (M = 2.74, SD = 1.76), t(196) = 3.69, p < .001.

6 The statistics here reflect only the ingroup addresses (the other two results concerning the outgroup or members external the organization are not reported here).

Chapter 4 51

Confidential Non-confidential

6

*** 5 4.00 4.00

4 3.44 *** 2.74

3

2 Intragroup motivations to pass on to Intragroupinformation pass motivations

1

0 Ingroup Outgroup

Group membership of information source

Figure 4. Study 1: Intragroup motivations to pass on information as a function of social identity and designated confidentiality.

Simple correlations The preceding results indicate that similar effects emerged for how useful the information was perceived to be (see Figure 3) and whether or not participants intended to share this information with ingroup members (see Figure 4). More specifically, non-confidential information was seen to be more useful (Mingroup = 4.22 vs. Moutgroup = 3.38) and, was more likely to be passed on to fellow ingroup members (Mingroup = 4.00 vs. Moutgroup = 2.74) when the information source was an ingroup rather than an outgroup member. Relatedly, confidential information was seen to be more useful (Mingroup = 3.81 vs. Moutgroup = 3.99) and was more likely to be passed onto fellow ingroup members (Moutgroup = 4.00 vs. Mingroup = 3.44) when it came from an outgroup rather than an ingroup member. However, note that these effects are stated in terms of means and do not reflect statistically significant differences. On this mean-difference basis though, we reviewed correlations to

Chapter 4 52 determine if perceived information utility acted as a potential mediator between social identity and confidentiality management. It was found that significant correlations emerged between perceived information utility and motivations to communicate the information. Importantly though, this was found to be true only in the case of non-confidential information for both ingroup, r(198) = 0.517, p < .001 and outgroup, r(198) = 0.458, p < .001. Critically then, in the case of confidential information, intentions to pass this type of information on to members of one’s own team did not depend on how useful the information was perceived to be for either the ingroup, r(198) = 0.107, p = 0.458 or the outgroup, r(198) = 0.212, p = 0.143. Discussion The focus for Study 1 was to examine how people respond to confidential information and how their reactions might differ as a function of the group membership of the source of information. In particular, we sought to determine if the extent to which people perceived information as confidential and what they intended to do with it depended on whether the information was provided by an ingroup or an outgroup member. The findings support H1 in showing that information was perceived to be more confidential when the source of that information was an ingroup member rather than an outgroup member irrespective of whether the information was described as 'general information' or 'confidential information'. Because shared social identity provides people with a common interpretative framework, this accelerates processes of communication, and thus should mean that when information is provided by a fellow ingroup member, one is more likely to potentially see it as ‘confidential’. On the other hand, we did not find support for H2a. That is, participants did not indicate that they were less likely to share confidential information than general information with another person (either ingroup, outgroup or outsize the organisation) if the information emerged from an ingroup member rather than an outgroup. In other words, people did not seem particularly motivated to preserve the confidentiality of material provided by an ingroup member. To account for this perplexing finding, it is noted that the use of imagined groups (as opposed to actual groups) might have not been robust enough methodologically to elicit a strong sense of identity (and thus limited our ability to find effects that may emerge from this identification). However, in support of H2b, individuals reported greater intentions to share 'general information' with ingroup members when the information emerged from an ingroup rather than an outgroup member. This makes sense if we recall that on the whole, common group membership is key for the success of our communication efforts. When identity is framed by a common sense of ‘we’, then individuals are not only more likely to pass on messages to members of their group but also to pay attention to the original message that is given. It follows then, that information described in more general terms is likely to be passed onto other ingroup members (assuming that it came

Chapter 4 53 from an ingroup member and thus deemed worthy of considering in the first place). Importantly too, we also explored other potentially relevant variables without specific hypotheses to uncover some evidence for process. At least four things are key to note here i) the group membership of the information source, ii) type of information, iii) perceived usefulness of the type of information provided and iv) intentions to further pass this information to other members of the ingroup. Accordingly, when an ingroup member provided non-confidential information, it was perceived to be more useful and more likely to be passed onto ingroup members. However, in the case of an outgroup information source, the same information needed to be labelled as confidential, for it to be considered useful and map onto sharing intentions. These effects were followed up with correlations to further unpack the processes that might dictate the flow of confidential information. In particular, it was found that although, perceived usefulness was positively correlated with intentions to pass on non-confidential (i.e., ‘general’) information, this relationship did not hold for intentions to pass on confidential information. From a SIT/SCT perspective, it could be argued that general information only seems useful when it comes from a favoured source (e.g., because ingroup members are perceived to share the same perspective on the world and thus their messages warrant attention and are trusted). However, in the context of non-shared group membership, perhaps the motivation to pass on confidential information of an outgroup source stems from a desire to undermine the group (along the lines of achieving a positive self-esteem) or to consolidate one’s status as a fellow group member (along the lines of sharing valuable information obtained from outside as a strategy to gain trust and respect). Summary Psychologists recently have become increasingly interested in the question of how individuals think about, and what they intend to do with, confidential information. Much of this research has taken an individual level of analysis. However, working from social identity principles, our study sheds light on possible psychological constructs that may influence how people manage confidential information across intergroup boundaries and within intragroup context. More particularly, this study provides evidence that social identity underpins perceptions of confidentiality. We also replicated a standard effect in the communication and social identity literature, that individuals share more general information with ingroup members than with outgroup members. There is also some evidence to support the claim that social identity plays a role in intentions to share confidential information. Although shared social identity does not directly translate into reduced intentions to share confidential information, non-shared social identity acts as a catalyst to share confidential information. Taken as a whole, the findings thus support claims that identity concerns structure how people think about confidential information but only lend indirect

Chapter 4 54 support to the idea that it is also structures what they intend to do with it.

Chapter 5 Study 2: Evidence that sharing confidential information creates shared social identity

Chapter 5 56 Overview This second study will seek to test the hypothesis that sharing confidential information can serve to build a sense of social identity between people in organizational contexts. Specifically, it will test whether the designated confidentiality of the stimulus material impacts on how the recipient of that information categorises themselves in terms of group identification. In this way, rather than examining how social identity is implicated in confidentiality management (as we did in Study 1), this study investigates the opposite side of the equation: whether confidentiality management is implicated in the development of social identity. More particularly, in this chapter I argue the case that, the exchange of confidential information does not necessarily constitute a communication “breach” but rather, can be a process through which people seek to build identity and consolidate their relationship with another. I also seek to establish what consequences (if any) flow on from this newly formed sense of identification with another. For this purpose, I will first provide a brief background to the study. This background is followed by an outline of key study hypotheses that are grounded in social identity theorizing. After that, I will detail the study methodology and findings. The chapter then concludes with a discussion of the implications of these findings before presenting conclusions from the research as a whole in the final chapter. Background Among scholars in psychology, the effects that emanate from the exchange of confidential information have historically received little systematic investigation. More particularly, the idea that some form of extra closeness can result when people share particular types of information is yet to be examined systematically. Nevertheless, I argued in Chapter 3 that there is a rich heritage to draw on from within the tradition of shared cognition to inform theorising around this issue. The key idea that emerges from this literature is that communication is a pivotal process through which individuals can develop a shared understanding— to the extent that it might also lead to a sense of shared identity (at the level of “us” and “what we represent”). Taken together, the research suggests that the communication process itself has important psychological consequences for identity-related outcomes (e.g., Koudernberg, et al., 2011). To explore this issue, the study in this second empirical chapter tests the following hypotheses derived from social identity theorizing.

H1: That when participants receive confidential rather than non-confidential or neutral information they will be less inclined to share this information with other ingroup members. Recall that one of the fundamental tenets of SCT is that the way in which an individual self- defines has important consequences for how they respond to their social world (Turner, 1982). More particularly, when the self is defined in terms of a social identity, then our internal dialogue

Chapter 5 57 becomes external, such that how we think or feel and determine what the appropriate course of action will be (in part) dependent on the shared group norms and values of the groups we belong to, and whose interests we seek to advance. In particular, because confidential information is typically sensitive in nature, then it is in the interests of the group member to ensure that the information remains confidential. From a SCT/SIT perspective, when a particular identity is salient, individuals should thus act in ways that will preserve the integrity of the group. Indeed, this is because the disclosure of confidential information is likely to compromise the group and its interests. In part too, the decision not to share confidential information reflects the fact that shared social identity provides individuals with the basis for a shared cognitive framework, such that any information that is offered is likely to be construed in the spirit in which it is given— so that if it is said to be confidential, it is treated as confidential.

H2: That when participants receive confidential rather than non-confidential or neutral information from an ingroup member this will increase their identification with the ingroup This hypothesis reflects the fact that receiving information that is sensitive and privileged is likely to increase the recipients’ sense that they are trusted by fellow group members (of whom the source is one), and therefore that they belong, and are seen to belong, in the group. This accords with research into gossip which suggests that this serves to create and consolidate intragroup bonds; e.g., Baumeister, Zhang & Vohs, 2004; McAndrew, Bell, & Garcia, 2007; Wert & Salovey, 2004).

H3: That when participants receive confidential rather than non-confidential or neutral information from an ingroup member, this will increase their sense of job satisfaction. This hypothesis follows from the leader-member exchange literature, which shows that the different ways in which leaders communicate with their followers can feed into levels of job satisfaction. In particular, studies show that due to resource constraint, only a select number of leader-follower dyads are characterised by high-quality exchanges, and that, when they occur, these promote mutual liking, trust, and respect and thus serve to enhance employee satisfaction and well- being (Loi, Chan & Lam, 2014; Loi, Mao & Ngo, 2009). Indeed, this logic might apply to confidential information which is by definition not widely shared and exclusive. As a corollary, though, employees who (for whatever reason) do not receive confidential information, are more likely to perceive that they are excluded from the ingroup and hence to disengage psychologically from the organizational unit and be less satisfied with their work.

Chapter 5 58 H4: That support for H3 will be mediated by H2. That is, when participants receive confidential rather than non-confidential information from an ingroup this will increase their sense of job satisfaction because it increases their sense of shared identity with that ingroup. This meditational hypothesis integrates H2 and H3 in suggesting that it is the increased sense of social identification that flows from the exchange of confidential information that accounts for increased job satisfaction on the part of recipients. More generally too, this hypothesis accords with a large body of organizational research which shows that social identification underpins a range of positive outcomes in the workplace (e.g., reduced stress, increased well-being; Haslam, 2004). The primary aim of this empirical study is to investigate whether, and to what extent the, exchange of confidential information is implicated in people’s development of a shared social identity. In this way, this study examines whether social identity might be enhanced by the communication of confidential information. More particularly, it seeks to examine the degree to which communication of ingroup information that is explicitly labelled as confidential rather than non-confidential or unlabeled builds a sense of shared identity. Because the manipulation of confidentiality did not work in the intended manner in the first study, it was decided a priori that in the present study, the only manipulation is that of the confidentiality label. In other words, the study that follows omitted the ingroup/outgroup manipulation and isolates the effect the confidentiality manipulation had solely. To achieve this, the study randomly assigned participants to one of three designated confidentiality conditions: confidential, neutral (i.e., unlabeled confidentiality) and non- confidential. Method Pre-testing Pre-test was conducted to check the robustness of the scales that were used in the study. It also sought to determine whether the manipulation of confidentiality worked in the intended manner. Twenty-three introductory psychology students took part in the pre-test. Of these, 13 (57%) were female and 10 (43%) were male. Their average age was 22 (SD = 3.20) ranging from 19 to 24 years. Participants were allocated to one of three conditions that specified the confidentiality of ingroup communications. It is important to note that group membership of the information provider was consistent across the three conditions of confidentiality. Here participants received in-group information that either was labelled as confidential, non-confidential or unlabeled (which served as 'control'). Respondents expressed their agreement with these statements (on scales from 1 = strongly disagree to 7 = strongly agree). ANOVA results indicated that participants perceived the condition described as

Chapter 5 59 'confidential' to be the most confidential (M = 6.62, SD =.96). Following this, the condition that represented the control level of confidentiality (unlabelled) was perceived to be of moderate confidentiality (M = 4.35, SD = 1.64). Finally, the condition described as ‘non-confidential’ was perceived to be the least confidential condition (M = 1.10, SD = 1.40). This pattern of results suggests that the manipulation of confidentiality was successful and, we therefore went on to use the scripts we had tested here for the main study. Participants and Design Three hundred and twenty-nine people were recruited through Amazon Mechanical Turk to participate in the study. This sample does not represent those that did not respond as instructed to the manipulation check or attention checks (73% response rate). Of the total sample size, 160 (49%) were female, and 169 (51%) were male. Their average age was 47 (SD = 5.13) ranging from 22 to 60 years. The majority of participants identified themselves as Australian (300, 91%) and indicated their ethnicity was Caucasian (256, 78%). Those wishing to participate in the study were required to provide their consent. The study design varied only one between-subjects factor: the designated confidentiality of the material provided to the participant. Here participants were randomly assigned to a condition that described information as either confidential or non-confidential, or (in a control condition) in which no label of confidentiality was provided. Perceived confidentiality, intentions to share information with ingroup, social identification, and perceived job satisfaction served as dependent measures. Participants were financially compensated for participating in the study and were debriefed upon its completion. The study was approved by the Ethical Review Committee in the School of Psychology at the University of Queensland (approval number: 15-PSYCH-MOP-04-JS). Procedure and Manipulations All participants were recruited through the online server Amazon Mechanical Turk. Individuals who agreed to participate were told that the purpose of the study was to examine attitudes and communication in the workplace. This served as a cover story. Participants were presented with a vignette that described an interaction in an unnamed organisation and imagine that they were part of this workplace, and on this basis, complete a self-report questionnaire. Participants were encouraged to pay close attention to the vignette and to the exchange of information that occurred within it. Participants were asked to imagine themselves in the study scenario as vividly as possible in order to be able to answer questions about their job position. The following information was provided to participants in the first part of the vignette:

Imagine you are working in an organisation. In this organisation, you work in Team Capricorn. You happen to be in the lunchroom alone, when Sam, a fellow member of your work team arrives. You notice

Chapter 5 60 he is carrying work files.

He turns to you and tells you about a memo he received from the head of operations.

He explains to you that the memo concerns another member of your work team, Taylor who has been unsuccessful for an internal promotion.

The confidentiality of the material given by the other employee was manipulated by providing participants with the following information:

The conversation between you and Sam flows for a while, and you briefly glance at the pile of documents Sam is carrying. You notice a print-out of an email with the following subject line:

CONFIDENTIAL: Taylor’s Application [in the explicit confidential condition]

NON-CONFIDENTIAL: Taylor’s Application [in the non-confidential condition]/

Taylor’s Application [in the control condition]

You and Sam continue talking about departmental issues, and he also mentions that the information he just gave to you about Taylor's application is [confidential/non-confidential/no mention of its inherent confidentiality]. You and Sam then leave the lunchroom together to return to work.

Note that aside from the manipulation of confidentiality, the content of the vignette was identical across the three conditions. After reading the vignette, participants answered a series of questions designed to measure their perceptions of, and intentions about how to handle, confidential information, as well as their identification and job satisfaction. These were as follows: Manipulation check. Perceived confidentiality. Perceived confidentiality was assessed using one item (“I believe the information the [colleague] gave to me was confidential”). Respondents indicated their agreement with this item on a scale from 1 = strongly disagree to 7 = strongly agree). Dependent variables. Intentions to share with ingroup. Three items measured motivations to communicate information received from an ingroup with other ingroup members ("I would share this information with people within in my work team", "I would be happy to let people within my work team know about this information" and "I would pass on the information Sam gave to me to people within my work team"). Respondents indicated their agreement with these statements (on

Chapter 5 61 scales from 1 = strongly disagree to 7 = strongly agree). The internal consistency of the overall scale was high, α = .96 and scores were therefore averaged to create a single measure. Shared identity. Shared social identification was rated by participants using Doosje, Ellemers and Spears’ (1995) measure. This measure includes four items that assesses cognitive, affective and evaluative dimensions of identification: "I see myself as a member of Team Capricorn", "I am pleased to be a member of Team Capricorn", "I feel strong ties with Team Capricorn" and "I identify with other members of Team Capricorn". Respondents indicated their agreement with these statements (on scales from 1 = strongly disagree to 7 = strongly agree). The internal consistency of the overall scale was acceptable, α = .81, and hence scores on the items were averaged to create a single measure. Perceived job satisfaction. Perceived job satisfaction was measured using the Michigan Organisational Assessment Questionnaire Job Satisfaction Subscale (MOAQ-JSS; Cammann, Fichman, Jenkins & Flesh, 1979). Scores on the MOAQ-JSS are computed using the mean of three items: "All in all I am satisfied with my job", "In general I like working here" and one reverse- scored item, "In general I don't like my job". In line with a recent meta-analysis (Bowling & Hammond, 2008), the internal consistency of the MOAQ-JSS was quite high, α = .83. Demographics. Participants were asked to indicate their age, gender, nationality, ethnicity and whether or not English was their first language. Age was treated as a continuous variable to preserve statistical power. Gender was coded as 0 = male and 1 = female. English as a first language was coded as 0 = yes and 1 = no. In total, participation the study took around 20 minutes to complete. After completing the questionnaire, participants were thanked for their participation and debriefed. Results Data analysis strategy. A series of ANOVAs were conducted to determine whether there was an effect of designated confidentiality on perceptions of confidentiality, shared identification and job satisfaction. To assess H5, a mediation model was tested using Hayes’ PROCESS macro in IBM Statistics (Hayes, 2013; Model 4). The current version of process macro does not permit the use of 3 levels within a categorical independent variable to test mediation. Thus, before testing the mediation model, the control condition of confidentiality (unlabelled) was removed (N=218). Manipulation check. The manipulation check appeared to be successful, such that there was a significant main effect of information type F(2, 326) = 279.24, p < .001, ηp²=.631 on perceived confidentiality which was decomposed by means of planned comparisons. These revealed that participants’ perceived information to be more confidential when the information was labelled confidential (M = 6.58, SD = .913) than when it was labelled non-confidential (M = 2.06, SD =

Chapter 5 62 1.83 , t(326) = 23.00, p < .001) or not labelled (neutral; M = 5.26, SD = 1.45, t(326) = 17.20, p < .001). These set of findings are presented in Figure 5 below.

***

9

6.58 8

5.26 7 ***

6

*** 5

4 2.06

3 Perceived confidentiality confidentiality Perceived 2

1

0 Non-Confidential Neutral Confidential

Designated confidentiality

Figure 5. Study 2: Perceived confidentiality as a function of designated confidentiality

Dependent variables. Intentions to share with ingroup. There was a significant main effect of information type F(2, 326) = 5.06, p < .007, ηp²=.030 on intentions to share information with the ingroup, which again was decomposed by means of planned comparisons. As predicted by H1, participants were generally less inclined to pass on information when the information was labelled confidential (M = 2.47, SD = 1.73) than when it was labelled as non-confidential (M = 3.14, SD = 1.70), t(326) = 2.77, p < .006. However, there was no difference in intentions to share with the ingroup when the information source labelled the information as non-confidential than when it was not labelled (M = 3.13, SD = 1.91), t(326) = - 0.66, p = 0.947. Shared identity. There was a significant main effect of information type F(2, 326) = 6.92, p

< .001, ηp²=.041 on participants’ identification with the ingroup. As predicted by H2, planned comparisons revealed that participants perceived a greater sense of shared identity with the ingroup when the information source labelled the information as confidential (M = 5.47, SD = 1.07) than

Chapter 5 63 when the information was labeled non-confidential (M = 4.89, SD = 1.28), t(326) = 3.66, p < .001, or when the information was not labelled, M = 5.27, SD = 1.14), t(326) = 2.35, p < .017. Job satisfaction. There was a significant main effect of information type, F(2, 326) = 3.08, p

< .05, ηp²=.01 which was again decomposed by means of planned comparisons. Consistent with H3, individuals reported greater job satisfaction when the information source labelled the information as confidential (M = 5.20, SD = 1.25) than when it was labeled non-confidential (M = 4.87, SD = 1.22) t(326) = -2.07, p < .03 or not labeled (neutral; M = 4.23, SD = 1.15), t(326) = 2.22, p = .026. Identification as a mediator. Bootstrapping analyses with 5000 resamples were conducted to test the indirect effects of confidentiality on perceived job satisfaction via social identification (Hayes, 2013; Model 4). Consistent with H4, the indirect effect of confidentiality via social identification was significant for perceived job satisfaction, b = 0.214, SE = .006 95% CI = .1054 to .3407. This is consistent with the claim that when participants received confidential (rather than non-confidential information) from an ingroup they reported higher levels of job satisfaction because they identified more strongly with the ingroup. Moreover, these results suggest that increases in social identity fully mediate the positive effect of receiving confidential (versus non- confidential) information on job satisfaction. This pattern of effects is displayed schematically in Figure 6.

Figure 6. Social identification mediates the relationship between designated confidentiality and job satisfaction.

Discussion The focus for this study was to explore how people react to receiving information at the level of the group and whether these responses differ as a function of information type. More particularly though, this study investigated and to what extent the, exchange of confidential information affects group identification. The above findings indicated that in support of H1, when participants received confidential rather than non-confidential or neutral information, they were less inclined to share this with other ingroup members. This is consistent with a key tenet of SCT/SIT

Chapter 5 64 that is, when a particular social identity is made salient, individuals will act in ways that seeks to protect (and advance) the interests of the particular group they belong to. In addition, the inclination to preserve the confidentiality of ingroup material may stem from the shared understanding that a sense of common identity provides. Consistent with H2, participants’ social identification with their team also varied as a function of the designated confidentiality of the material they received, such that those who received confidential information reported a stronger sense of identification with their ingroup than those who received non-confidential or unlabelled information. This finding is in line with the research on gossip, in that it reflects the role certain information can play in consolidating intragroup bonds. This strengthening of ingroup identification may occur because the recipient perceives that they have been trusted with confidential information that belongs to their group. Similarly, consistent with H3, participants reported higher levels of job satisfaction when they received confidential (rather than non-confidential or unlabelled) information. This suggests that the decisions individuals’ make with the passing on of confidential information bears the potential to create a distinct ingroup/outgroup dynamic. Moreover, such decisions can influence the ways in which employees think about their life at work. Indeed, this aligns with the leader-member exchange literature that shows us that high-quality exchanges, in particular, are also exclusive and tend to be associated with higher levels of job satisfaction. Finally, too, in line with H4, path analysis indicated that support for H3 was fully mediated by support for H2. In other words, it appears that participants who received confidential rather than non-confidential or unlabeled information from an ingroup member reported a higher job satisfaction because receiving this information increased their identification with their ingroup. This finding is consistent with a wider body of work that demonstrates the ways in which being rooted in a social network (and feeling part of that network) yields authentic psychological benefits at both personal and social levels (e.g., Haslam, 2004). Summary Research from different disciplinary domains and more particularly, the emerging literature on shared cognition points to the idea that social identity can be encouraged by particular patterns of communication and that this, in turn, has a range of positive consequences. This study is the first of its kind, to test— and find support for— the idea that ostensibly confidential information, in particular, influences self-categorization on the part of the recipient of that information. In particular, this leads us to speculate that the exchange of confidential information might serve as a social cue to the recipient regarding their position within the ingroup. Another fundamental message that emerges from this work is that shared social identification, does indeed serve as a conduit for the fate of confidential information within the

Chapter 5 65 workpla ce. For we can surmise that it was only because the recipient categorised themselves as an ingroup member, that they were inclined to keep confidential information as confidential. Although the study would need to have included conditions from an outgroup source in order to have strong support for this claim, this leads us to suspect that the confidentiality of the material is respected only within the context of shared group membership. In short, it is shared (rather than non-shared) social identity that acts as a catalyst for preserving the confidentiality of ingroup material. In addition, another key point that emerges from this study is that individuals are more likely to feel good about their work in the context of confidential information exchange because such makes individuals’ identify more strongly with the team. Taken as a whole, this set of findings speak to the idea that as well as building identification, the exchange of confidential information can have important flow on effects for an organization as a whole. Moreover, while these effects are generally understood to be negative (because they breach organisational rules), it is clear that they can also be positive. This in turn suggests that the motivations underpinning breaches of confidentiality are likely to be ‘mixed’, rather than only ever negative.

Chapter 6 General discussion and conclusions

Chapter 6 67 Introduction and Overview This thesis sought to examine the factors that determine how individuals manage confidential information in organizational contexts, and the effects of managing confidential information in particular ways. Although the literature has seen a growth of interest in issues of confidentiality management, progress in the field had been stifled by the lack of a theoretical framework that could account for its group-based underpinnings. I argued that prevailing approaches are limited in scope and do little to help us understand the exchange or interpretation of confidential information because they do not account for the identity-based relationships that underpin those dynamics. More particularly, I argued that social identity theory (Tajfel & Turner, 1979 and Turner, 1985) and social categorization theory (Turner, 1985; Turner, et al., 1987) are ideal theoretical frameworks for understanding the social dynamics involved in management of confidential information. As such, this framework is a solid first step in creating a more inclusive and clearer picture of the dynamics between social relationships and the management of confidential information. The empirical studies presented in this thesis were designed to examine two sides of the same process. On the one side, they examined the ways in which issues related to group membership and social identity are implicated in the management of confidential information (e.g., to perceive information to be ‘confidential' and be willing to preserve, or else violate confidentiality). On the other side, they examined the ways in which the exchange of confidential information is implicated in people's development of a shared social identity. Taken together, these studies tested the proposition that social identity both affects, and is affected by, the exchange (or non-exchange) of confidential information. Across the two studies, I provided evidence that is broadly consistent with this proposition. In this concluding chapter, I will review the respective aims of each of the empirical studies and outline key findings that stem from this work. I will also focus on how these two studies enhance our understanding of how people think about confidential information and what they intend to do with it, as well as the psychological impact of confidential information exchange. I will then discuss broad implications of this research, as well as its limitations, before outlining potential avenues for research and providing some concluding remarks. Social Identity as a Precursor to Confidentiality Management The primary aim of Study 1 was to establish the extent to which an informational source’s group membership (ingroup vs. outgroup) impacts the information recipient’s perceptions of confidentiality and their motivations to communicate confidential information in an organizational setting. As predicted by H1, one key result to emerge from this study was that perceptions of confidentiality depend on shared group membership. Specifically, when an ingroup member (rather than an outgroup member) provided any type of information (confidential or non-confidential),

Chapter 6 68 individuals consistently perceived that information to be more confidential. This finding accords with one key tenet of SCT, namely, that shared social identity is the basis for social influence (Turner, 1991). Thus, when people perceive each other as members of the same salient social category in a given context, they are motivated, and exert considerable effort, to reach agreement with them (Haslam et al., 2012). This means that people are more inclined to treat sensitive information provided by ingroup members as confidential. In part too, this inclination to see potentially confidential as ‘confidential’ will reflect that a sense of shared social identity enables individuals to trust, communicate and respect one another (along the lines suggested by Tyler & Blader, 2000). In this way, all these elements that underpin social relationships are also critical for the success of our communication efforts — including those that extend to confidential information exchange (Haslam et al., 2012) However, an important caveat to this finding was that, and rather surprisingly, regardless of the group membership of the information provider, the perceived confidentiality of information did not increase when the provider explicitly labeled it as confidential. Indeed, on the contrary, participants perceived information labeled confidential to be less confidential than non-confidential information. One tentative explanation for this is that people questioned how confidential the information really was, simply by virtue of the fact that it has been shared. Another possibility was that people were unsure of whether the non-confidential information was confidential (because it is labeled as ‘general’) and therefore tend to err on the side of caution. In terms of what people intended to do with the information, a complicated set of findings emerged from this study. Contrary to H2a, intentions to share confidential information with another person (ingroup, outgroup or outside the organization) did not differ as a function of the group membership of the information source. In other words, it did not matter if the confidential information emerged from an ingroup or an outgroup member, participants indicated that they would be equally likely to share it. This is a perplexing finding in that SCT/SIT perspective might lead us to speculate that giving away confidential information of the ingroup might not reflect a means of advancing the interests of the group and/or might be perceived to constitute a violation of one’s group membership status. What accounts for this unexpected finding? It is possible that a heightened respect for ingroup members’ confidentiality may only emerge for participants who identify strongly with the ingroup. Here, we did not differentiate between people with high and low levels of identification with the ingroup. Thus, in order to clarify our understanding, in future research we need to determine whether the strength of individuals' group identity moderates their willingness to respect the confidentiality of information provided by ingroup member. This issue might be further compounded by the fact that the study used imagined groups presented in a vignette, rather than

Chapter 6 69 actual groups, which may not have elicited a strong sense of identity to begin with. Nevertheless, in support of H2b, findings indicated that people were more inclined to share general information with fellow ingroup members when it emerged from an ingroup rather than an outgroup member. In particular, this occurs here because people will see themselves as sharing the same perspective on the world and the same interests (Turner, 1991). The first part of this finding replicates the standard effect in the literature, that generally people are motivated to share information with people whom they perceive to be members of the same salient social category (i.e., ingroup rather than outgroup members) (e.g., Agama, 1997; Morton et al., 2012, Suzuki, 1998). Moreover, though, the second part of this finding is in line with a wider body of research that shows us that common group membership underpins the messages we pay attention to and how we interpret them (e.g., Hornsey & Imani, 2004; Mackie et al., 1990). Interestingly, findings also indicated that an ingroup source elicited a tendency to share general information rather than confidential information but an outgroup source elicited the opposite effect. More particularly, if the outgroup provided confidential information, people were more willing to break confidentiality and inform members of their ingroup, even after being explicitly told that the information given to them was confidential. This set of findings together might not translate into a direct willingness to preserve the confidentiality of ingroup material, but do, indicate at the very least that, an absence of shared social identity dictates a refusal to preserving confidentiality. As such, the original standard main effect in the wider literature on social identity and communication is reversed. So although people are generally motivated to share information with ingroup members if it emerges from an ingroup member, our analysis suggests that confidential information starts to be spread around if it emerges from an outgroup member. In line with social identity theorising, a range of strategic considerations might dictate why this might be the case. In particular, perhaps sharing an outgroup member’s confidential information is a means of empowering the ingroup and exerting more influence. Alternatively, it may represent an attempt to convey the ingroup’s superiority over relevant outgroups (van Leeuwen, & Täuber, 2010). In addition, when people come to identify the victories and disappointments of a particular organizational unit as their own, they might be more likely to undermine the outgroup in order to advance the interests of their own group (e.g., Mael & Ashforth, 1992). Relatedly, preserving the confidentiality of information provided by ingroup members may be underpinned by concern for the welfare of the ingroup, whereas breaching confidentiality of information provided by the outgroup is likely to be based on a strategic analysis of costs and benefits (Stürmer & Snyder, 2009). In addition to this social identity process, perceived information utility also determined receiver’s intentions to share information further with others. That is, participants perceived non-

Chapter 6 70 confidential information to be more useful when the information provider was an ingroup member and were more inclined to share this information with ingroup members. Critically, though, this pattern was not present for confidential information. Here there was no significant difference between intentions to share confidential information provided by an outgroup member versus an ingroup member, and perceived information utility was not significantly related to intentions to share. This may be because confidential information was perceived to be more useful in general and overshadowed the impact of the source’s group membership. Another possibility is that although perceived information utility influences sharing of non-confidential information, another psychological construct underpins the containment of confidential information. Thus, a number of key questions arise from this analysis. The first is that, why are people motivated to preserve the confidentiality of ingroup information? The second asks, why are people motivated to undermine the confidentiality of outgroup information? Finally, does the strength of group identity play a role in motivations to preserve confidential material from the ingroup? Although it is only a preliminary study, our study leads us to speculate on possible psychological constructs that may govern how people think about confidential information as well as the conditions under which they might preserve or else, violate confidentiality.

Social Identity as a Product of Confidentiality Management The primary aim for Study 2 was to determine whether sharing confidential information can serve to build a sense of social identity between people in organizational contexts. Specifically, we sought to systematically examine whether the designated confidentiality of the stimulus material impacts on how the recipient of that information categorises themselves in terms of group identification. In addition to this, I sought to establish if any effects emerged on the basis of this identification. However, before moving onto the findings of these hypotheses, the hypothesis that follows is not only common to both studies, but is the most straightforward one from which to begin. As predicted by H1, when participants received confidential rather than non-confidential or unlabeled information from an ingroup information source, they indicated that they were less inclined to share this with other ingroup members. This finding is in accordance with one of the fundamental tenets of SCT, in that when a particular identity is salient, the process of seeking to serve to protect (and advance) the interests of their psychological group will motivate them to preserve the confidentiality of ingroup material. In these specific circumstances, they will want to ensure confidential information is contained as much as possible because in doing so, they are, actually promoting the interests of the group (Turner et al., 1987; Haslam et al., 2012). Moreover, because common group membership underpins, and facilitates, a mutual understanding between

Chapter 6 71 parties to the communication process, then this should mean that confidential information will be construed, and treated in the way that it was intended. As such, this latter finding aligns with other evidence that shared social identity impacts on behaviour via motivational and cognitive channels (Brewer, 1979; Spears & Otten, 2012). The most fundamental point to emerge from this finding is that confidentiality is likely to be preserved in the context of shared group membership because here people are motivated to respect its confidential status. As a corollary, we expect that they may be less likely to protect confidential information provided by an outgroup, although it should be noted that this was not a hypothesis that we examined in Study 2. In support of H2, we also found that when participants received confidential rather than non-confidential or unlabeled information from an ingroup member, this increased their sense of identification with that ingroup. In general terms, this finding is consistent with literature in the shared cognition tradition that speaks to the capacity for a social identity to develop when a shared understanding emerges among group members through high-quality communication (e.g., Haslam, Turner, Oakes, Reynolds & Doosje, 2002; Koudenburg, et al., 2013). More particularly though, it is argued that confidential communication accelerates this process for a sense of self to emerge among group members because distinctive features of confidentiality— in particular, its capacity to communicate respect and trust — help to promote social bonds (van Dick and Haslam, 2012). This finding is consistent with a wide body of research on gossip that also argues for its powerful role in psychologically binding individuals to one another on an interpersonal plane (e.g., Baumeister et al., 2004; McAndrew et al., 2007). Findings also indicated support for H3, in that when individuals received confidential rather than non-confidential or unlabelled information from an ingroup member, they reported higher levels of job satisfaction. This finding aligns with studies that show that only a select few leader-follower dyads characterised by mutual liking, trust, and respect tend to reap positive emotional outcomes for employees (Amos & Herrick, 2005; Gerstner & Day, 1997). In this respect, given that confidential information tends to be exclusive, then its communication (or lack thereof) will also create a distinctive ingroup/outgroup state of affairs that may spill over into how employees evaluate their job. Specifically, employees who receive confidential information might perceive that they are valued by their ingroup and feel ‘good’ about being included and hence engage psychologically with the organisational unit and are satisfied with their work. On the other hand, employees who do not receive confidential information might perceive that they are excluded from the ingroup, which might heighten disengagement and lead to lower levels of job satisfaction. Consistent with the mediational hypothesis of H4, findings indicated that it was the increased sense of social identification that flows from the exchange of confidential information that accounts for increased job satisfaction on the part of recipients. This notion is consistent with

Chapter 6 72 an extensive body of organisational research that shows us the ways in which social identity principles have facilitated our understanding of a number of organisational outcomes at both personal and social levels (Haslam et al., 2005; see also Mael & Ashforth, 1995; Riketta, 2005). Despite the traditional view in the confidentiality literature that breaches of confidentiality are isolated incidents contingent upon individual decisions, our research establishes that these exchanges also have important consequences for the social structure of relationships. In particular, this work speaks to the notion that confidentiality management has a potential social function in fostering group identification by satisfying a fundamental need to belong. In other words, although sharing confidential information has stereotypically collected a poor reputation, it is also a vehicle through which we create important connections to one another and gain key insight into our relative position within the group, and thus might function as an instrument of social order (along the lines suggested by the gossip literature e.g., Baumeister et al., 2004). In addition, the exchange of confidential information can be seen to be related to other constructs central to aspects of social relationships such as trust-building. This means that by building identification, the sharing of confidential information can give rise to positive work outcomes for both the individual and the organisation in which they are embedded. Broader Implications Taken together, this work has important implications for the literature on confidentiality management. Although scholarship to date has made considerable steps forward in understanding this topic, the field as it stands is hampered by its exclusive focus on individuals as individuals. In this regard, our various findings speak, first, to the idea that the social self is an important and integral part of how issues of confidentiality play out. Second, although individuals involved in breaches of confidentiality have typically been judged poorly, our findings suggest that this act may actually be a tool to promote social order and structure. This is because the person receiving confidential information may perceive it to be a sign of deep trust such that it, thereby is a means for increasing their standing within a group and thereby their identification. In this respect, confidential information is a resource –– one that can be mobilised to build identification. Related to this, we see that confidential information is a potent form of currency. Indeed, because it is information that has a particular value, it can be used to build social capital. From a practical perspective, a question remains as to how organisations can direct the behaviour of their employees to protect confidential company information. Because leaks are associated with a significant financial cost (Pricewaterhouse & Coopers, 2002), it makes sense that a desire to circumvent leaks is a high organisational priority. Specifically, research suggests that organisations tend to take a preventative approach by limiting field-testing of products or restricting

Chapter 6 73 which employees are permitted to have certain information (e.g., see Nocera, 2008, see also, Lashinsky, 2012, on the lengths Apple Inc goes to protect commercial information). However, our work is a solid first step in arguing that a shared sense of group identification between people heightens the possibility that they will understand information to be confidential and act on it accordingly, and that it is under these circumstances that we are most likely to see effective management of confidential information. Thus, it may be possible to promote optimal outcomes in the management of confidential information by also working to boost individuals’ sense that they are part of a relevant organizational unit (producing a sense that ‘we are in this together’). Limitations and Future Directions By applying social identity theorising to confidentiality management as well considering both bottom up and top down processes we laid the groundwork for future researchers to disentangle the social mechanisms that impact confidentiality management. Importantly too, the experimental studies ensure key factors that might otherwise contribute to decisions regarding confidentiality in the professional sphere, such as past experience and nature of relationship/s, are held constant. More generally, longitudinal and prospective research is clearly needed to examine the interplay of effects between group identification and social identity. Although the current study has provided initial evidence for the idea that social identity and confidentiality management are empirically linked, there are some limitations inherent in the experiments. Although we found evidence consistent with our prediction that information is more likely to be construed as 'confidential' in the context of shared group membership, the failure of our experimental manipulation of confidentiality in Study 1 meant that the role of group membership in the observed effects was unexpectedly complex. To overcome this limitation as well as ensure a more robust manipulation, the follow-up study not only omitted the ingroup/outgroup manipulation to isolate the single effect of confidentiality but also included a more structured and objective reference to the inherent confidentiality of the stimulus material. This manipulation proved successful, and the results within the second study illuminated a far more compelling set of effects. More particularly, the idea that individuals are more motivated to preserve the confidentiality of ingroup material in the context of shared group membership was empirically supported here and thus substantiated theoretical claims. However, neither study specified the conditions responsible for the different outcomes of confidentiality. That is, what are the variables the moderate the effect of social identity on confidentiality management (e.g., why are people motivated to preserve the confidentiality of ingroup information or else, undermine the confidentiality of outgroup information). From a SCT/SIT perspective, we might speculate that in the case of non-confidential information, people will be motivated to transfer such information on the basis of perceived utility. However, in the case

Chapter 6 74 of confidential information, motivations here cease to be contingent on the value or usefulness of the information. Instead, people will indicate a willingness to ensure ingroup information is kept confidential in the context of shared group membership on the basis of respect for the ingroup member. Simultaneously, they will intend to broaden the boundaries of outgroup confidential information because what matters here is disrespect. Clearly, the former situation is conducive to preserving confidentiality in a way that the latter is not. This represents one avenue for future research to consider. In presenting our studies thus far, an important proviso is that the risk of not preserving confidentiality for the outgroup may be overcome if the salient social identity in question is (re)defined superordinately. More specifically, at one level of abstraction the employee of an opposing work-team would be construed as an outgroup member but at a higher level the employee will be recategorized as an ingroup member subsumed under the overarching sense of organizational identity. Therefore, future research should also examine the impact of categorisation of this form on confidentiality management. In a related vein, the empirical studies within this thesis are limited in that they only focus on participants’ responses to trivial groups, which do little to clarify our understanding in personally relevant situations. The rational sceptic might argue that the results may not reflect how individuals encounter or deal with issues of confidentiality in organizational settings where such decisions matter. Such findings are likely to shed light on the challenges people confront in dealing with issues of confidentiality. That is, it might be important to elicit the processes antecedent to breaches of confidentiality and consequences of such breaches. Only then, is one in a better position to specify what conditions are responsible for the different outcomes of confidential information. In doing so, future research might not only be in a position to achieve theoretical progress on this important topic, but also to triangulate these preliminary findings (Denzin & Lincoln, 2005). Concluding Comment The overarching goal of this thesis was to explore the social psychology of confidentiality management. I drew on theorising in the social identity tradition to understand how people manage confidential information, as well as the effects that emerge from the exchange of confidential information within the organisational domain. In particular, it examined whether social identity might function as a precursor and a product of the way in which confidential information is managed. The findings confirmed our predictions that shared social identity influences (in complex ways) people’s management of confidential information, and also showed that the act of sharing confidential information can serve to build a sense of social identity. This meant that people were more likely perceive information as ‘confidential' in the context of shared group membership, and

Chapter 6 75 relatedly, people were more likely to develop a sense of organizational identification after receiving confidential (rather than non-confidential) information from a colleague. Rather than processes of confidentiality management being understood at the individual level, the present research shows that how people manage confidential information in organisational settings is contingent upon social relationships. This, in turn, has a range of practical implications not only for the way we think about why people might breach confidentiality but also for the way in which organisations might educate and train their staff. In particular, this research leads us to speculate that issues of confidentiality might be better understood and tackled through greater appreciation of the importance of shared identity. Our sense of identity (and the factors concomitant to that identity; such as trust, open communication, support) may act as a critical determinant in the way in which people manage confidential information at work.

References Adler, R. (1995). To tell or not to tell: the psychiatrist and child abuse. Australian and New Zealand Journal of Psychiatry, 29, 190 – 198. Agama, A. (1997). The communication of information in an organizational setting: The role of self- categorization and perceived group membership. PhD diss., Australian National University. Allen, A. (2011). Privacy and medicine. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy. Stanford, CA: Stanford University. Allen, A. L. (2008). Confidentiality: an expectation in health care. Faculty Scholarship. Paper 186. Amiot, C. E., De la Sablonniere, R., Terry, D. J., & Smith, J. R. (2007). Integration of social identities in the self: Toward a cognitive-developmental model. Personality and Social Psychology Review, 11(4), 364-388. Amos, M. A., Hu, J., & Herrick, C. A. (2005). The impact of team building on communication and job satisfaction of nursing staff. Journal for Nurses in Professional Development, 21(1), 10-16. Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organization. Academy of Management Review, 14(1), 20-39. Australian Institute of Health and Welfare. Australia's Welfare 2003: The sixth biennial welfare report of the Australian Institute of Health and Welfare. Canberra Australia: Australian Institute of Health and Welfare; 2003. AIHW Cat No AUS 41. pp 65-158. Baird, K. A., & Rupert, P. A. (1987). Clinical management of confidentiality: A survey of psychologists in seven states. Professional Psychology: Research and Practice, 18, 347–352. Baker, J. S., & Jones, M. A. (1996). The poison grapevine: How destructive are gossip and rumor in the workplace? Human Resource Development Quarterly, 7(1), 75-86. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: desire for interpersonal attachments as a fundamental human motivation. Psychological bulletin, 117(3), 497. Beauchamp, T. L., & Childress, J. F. (1994). Principles of medical ethics. Oxford University Press, USA. Beauchamp, T. L., & Childress, J. F. (2001). Principles of biomedical ethics. Oxford University Press, USA. Bersoff, D. N. (2008). Ethical conflicts in psychology (4th ed.). Washington, DC: American Psychological Association. Boroş, S., Curşeu, P. L., & Miclea, M. (2011). Integrative tests of a multidimensional model of organizational identification. Social Psychology. Bourhis, R. Y. (1991). Organizational communication and accommodation: Toward some conceptual and empirical links. Contexts of accommodation: Developments in applied sociolinguistics, 270-303.

77 Bowling, N. A., & Hammond, G. D. (2008). A meta-analytic examination of the construct validity of the Michigan Organizational Assessment Questionnaire Job Satisfaction Subscale. Journal of Vocational Behavior, 73(1), 63-77. Brann, M., & Mattson, M. (2004). Toward a typology of confidentiality breaches in health care communication: An ethic of care analysis of provider practices and patient perceptions. Health Communication, 16(2), 231-251. Brewer, M. B. (1979). Ingroup bias in the minimal intergroup situation: A cognitive-motivational analysis. Psychological Bulletin, 86, 307–324. Cain, P. (1998). The limits of confidentiality. Nursing Ethics, 5, 158–165. Cammann, C., Fichman, M., Jenkins, D., & Klesh, J. R. (1983). Assessing the attitudes and perceptions of organizational members. Assessing organizational change: A guide to methods, measures, and practices, 71, 138-141. Cannon-Bowers, J. A., & Salas, E. (2001). Reflections on shared cognition. Journal of Organizational Behavior, 22(2), 195-202. Carlisle, J., Shickle, D., Cork, M., & McDonagh, A. (2006). Concerns over confidentiality may deter adolescents from consulting their doctors. A qualitative exploration. Journal of Medical Ethics, 32(3), 133-137. Caughlin, J. P., & Petronio, S. (2004). Privacy in families. Handbook of family communication, 379-412. Cheng, T. L., Savageau, J. A., Sattler, A. L., & DeWitt, T. G. (1993). Confidentiality in health care: a survey of knowledge, perceptions, and attitudes among high school students. JAMA, 269(11), 1404-1407. Clark, C. (2006). Against confidentiality? Privacy, safety and the public good in professional communications. Journal of Social Work, 6(2), 117–136. Clark, P. G., Cott, C., & Drinka, T. K. (2007). Theory and practice in interprofessional ethics: A framework for understanding ethical issues in health care teams. Journal of Interprofessional Care, 21, 591–603. Cline, R. J., & McKenzie, N. J. (2000). Dilemmas in disclosure in the age of HIV/AIDS: Balancing privacy and protection in the health care context. In S. Petronio (Ed.), Balancing the secrets of private disclosures (pp. 71–82). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Coopers, P. W. (2002). Innovation in the Australian building and Construction Industry–Survey Report. In Australian Construction Industry forum (ACIF). Cordess, C. (ed.) (2000) Confidentiality and Mental Health. London: Jessica Kingsley.

78 Cummings, N. A. (2000). A psychologist’s proactive guide to mental health care: New roles and opportunities. In A. Kent & M. Hersen (Eds.), A psychologist’s guide to managed mental health care (pp. 141–161). Mahwah, NJ: Erlbaum. Davidson, J. R., & Davidson, T. (1996). Confidentiality and managed care: Ethical and legal concerns. Health & Social Work, 21(3), 208–215. Davis, J. L., & Mickelson, D. J. (1994). School counselors: Are you aware of ethical and legal aspects of counseling? The School Counselor, 42(1), 5-13. Derlega, V. J., Metts, S., Petronio, S., & Margulis, S. T. (1993). Self-disclosure. Newbury Park, CA: Sage. Denzin, N. K., & Lincoln, Y. S. (2005). The SAGE handbook of qualitative research (3rd ed.). Thousand Oaks, CA: Sage. Doosje, B., Ellemers, N., & Spears, R. (1995). Perceived intragroup variability as a function of group status and identification. Journal of Experimental Social Psychology, 31(5), 410-436. Dorodny, V. S. (1998). Security and confidentiality of health information systems: Implications for physicians. Physician Executive, 24(3), 32–37. Dubois, M. (Ed.). (2002). Psychologists and pain physicians: Can they share all the information about their patients? Pain Medicine, 3, 169–171. Duck, J. M., & Fielding, K. S. (1999). Leaders and subgroups: One of us or one of them? Group Processes & Intergroup Relations, 2(3), 203-230. Duncan, R. E., Williams, B. J., & Knowles, A. (2012). Breaching confidentiality with adolescent clients: A survey of Australian psychologists about the considerations that influence their decisions. Psychiatry, Psychology and Law, 19(2), 209–220. Dunbar, R. (1996). Gossip, grooming and the evolution of language. Cambridge: Harvard UP. Dyer, J.H. and Nobeoka, K. (2000), “Creating and managing a high-performance knowledge- sharing network: the Toyota case”, Strategic Management Journal, Vol. 21 No. 3, pp. 345-67. Elger, B. S. (2009). Factors influencing attitudes towards medical confidentiality among Swiss physicians. Journal of Medical Ethics, 35(8), 517–524. Ellemers, N., De Gilder, D., & Haslam, S. A. (2004). Motivating individuals and groups at work: A social identity perspective on leadership and group performance. Academy of Management Review, 29(3), 459-478. Ellemers, N., Spears, R., & Doosje, B. (2002). Self and social identity. Annual Review of Psychology, 53(1), 161-186. Fennig, S., Barak, V., Benyakar, M., Farina, J., Blum, A., & Treves, I. (2000). Comparison of the attitudes of Israeli therapists and lay persons to ethical dilemmas in psychotherapy. Journal of Nervous and Mental Disease, 188(11), 777-785.

79 Fisher, M. A. (2008). Protecting confidentiality rights: The need for an ethical practice model. American Psychologist, 63(1), 1. Foddy, M., Platow, M. J., & Yamagishi, T. (2009). Group-based trust in strangers the role of stereotypes and expectations. Psychological Science, 20(4), 419-422. Gaertner, L., Iuzzini, J., Witt, M. G., & Oriña, M. M. (2006). Us without them: evidence for an intragroup origin of positive in-group regard. Journal of Personality and Social Psychology, 90(3), 426. Gaertner, S. L., Mann, J., Murrell, A., & Dovidio, J. F. (1989). Reducing intergroup bias: The benefits of recategorization. Journal of Personality and Social psychology, 57(2), 239. Gardner, M. J., Paulsen, N., Gallois, C., Callan, V. J., & Monaghan, P. (2001). Communication in organizations: An intergroup perspective. The new handbook of language and social psychology, 2, 561Á584. Gerstner, C. R., & Day, D. V. (1997). Meta-Analytic review of leader–member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82(6), 827. Giles, H. (2008). Communication accommodation theory. Sage Publications, Inc. Chicago. Gostin, L. O., Hodge Jr, J. G., & Burghardt, M. S. (2002). Balancing Communal Goods and Personal Privacy Under a National Health Informational Privacy Rule. St. Louis LJ, 46, 5-1091. Graen, G. (1976). Role-making processes within complex organizations. Handbook of industrial and organizational psychology, 1201, 1245. Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The Leadership Quarterly, 6(2), 219-247. Graen, G., & Cashman, J. F. (1975). A role-making model of leadership in formal organizations: A developmental approach. Leadership frontiers, 143, 165. Greenaway, K. H., Wright, R. G., Willingham, J., Reynolds, K. J., & Haslam, S. A. (2015). Shared identity is key to effective communication. Personality and Social Psychology Bulletin, 41(2), 171-182. Hadnagy, C. (2010). Social engineering: The art of human hacking. John Wiley & Sons. Hale, A., & Borys, D. (2013). Working to rule, or working safely? Part 1: A state of the art review. Safety Science, 55, 207-221. Harwood, I. A., Turnock, S. R., & Ashleigh, M. J. (2014). When bubbles burst: Mimetic insights into minimizing confidentiality breaches. European Management Journal, 32, 84–90. Haslam, S. A. (2001). Psychology in organizations. Sage. Haslam, C., Morton, T. A., Haslam, S. A., Varnes, L., Graham, R. & Gamaz, L. (2012). “When the age is in, the wit is out”: Age-related self-categorization and deficit expectations reduce

80 performance on clinical tests used in dementia assessment. Psychology and Aging, 27(3), 778- 784. Haslam, S. A. Eggins, R. A., & Reynolds, K. J. (2003). The ASPIRe model: Actualizing Social and Personal Identity Resources to enhance organizational outcomes. Journal of Occupational and Organizational Psychology, 76(1), 83-113. Haslam, S. A., & Platow, M. J. (2001). The link between leadership and followership: How affirming social identity translates vision into action. Personality and Social Psychology Bulletin, 27(11), 1469-1479. Haslam, S. A., Jetten, J., O'Brien, A., & Jacobs, E. (2004). Social identity, social influence and reactions to potentially stressful tasks: support for the self-categorization model of stress. Stress and Health, 20(1), 3-9. Haslam, S. A., O'Brien, A., Jetten, J., Vormedal, K., & Penna, S. (2005). Taking the strain: Social identity, social support, and the experience of stress. British Journal of Social Psychology, 44(3), 355-370. Haslam, S. A., Powell, C., & Turner, J. (2000). Social identity, self-categorization, and work motivation: rethinking the contribution of the group to positive and sustainable organisational outcomes. Applied Psychology, 49(3), 319-339. Haslam, S. A., Reicher, S. D., & Levine, M. (2012). When other people are heaven, when other people are hell: How social identity determines the nature and impact of social support. The social cure: Identity, health and well-being, 157-174. Haslam, S. A., Turner, J. C., Oakes, P. J., Reynolds, K. J., & Doosje, B. (2002). From personal pictures in the head to collective tools in the world: How shared stereotypes allow groups to represent and change social reality. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Press. Hogg, M. A. (2000). Subjective uncertainty reduction through self-categorization: A motivational theory of social identity processes. European Review of Social Psychology, 11(1), 223-255. Hogg, M. A., & Terry, D. I. (2000). Social identity and self-categorization processes in organizational contexts. Academy of Management Review, 25(1), 121-140. Hook, M. K., & Cleveland, J. L. (1999). To tell or not to tell: Breaching confidentiality with clients with HIV and AIDS. Ethics & behavior, 9(4), 365-381. Hornsey, M. J., & Imani, A. (2004). Criticizing groups from the inside and the outside: An identity perspective on the intergroup sensitivity effect. Personality and Social Psychology Bulletin, 30(3), 365-383. Hornsey, M. J., Oppes, T., & Svensson, A. (2002). “It's OK if we say it, but you can't”: responses to

81 intergroup and intragroup criticism. European Journal of Social Psychology, 32(3), 293-307. Hughes, J. C. and Louw, S. J. (2002) ‘Confidentiality and Cognitive Impairment: Professional and Philosophical Ethics’, Age and Ageing 31(2): 147–50. Isaacs, M. L., & Stone, C. (2001). Confidentiality with minors: Mental health counselors’ attitudes toward breaching or preserving confidentiality. Journal of Mental Health Counseling, 23(4), 342–356. Jacob-Timm, S. (1999). Ethically challenging situations encountered by school psychologists. Psychology in the Schools, 36(3), 205-217. James, S. D., Leiblein, M. J., & Lu, S. (2013). How firms capture value from their innovations. Journal of Management, 39(5), 1123-1155. Jenkins, G., Merz, J. F., & Sankar, P. (2005). A qualitative study of women’s views on medical confidentiality. Journal of Medical Ethics, 31(9) 753–759. Jetten, J., Haslam, C., & Alexander, S. H. (Eds.). (2012). The social cure: Identity, health and well- being. Psychology Press. Jones, E., Watson, B., Gardner, J., & Gallois, C. (2004). Organizational communication: Challenges for the new century. Journal of Communication, 54(4), 722-750. Kämpf, A., & McSherry, B. (2006). Confidentiality in therapeutic relationships: The need to develop comprehensive guidelines for mental health professionals. Psychiatry, Psychology and Law, 13(1), 124–131. Kämpf, A., McSherry, B., Thomas, S., & Abrahams, H. (2008). Psychologists’ perceptions of legal and ethical requirements for breaching confidentiality. Australian Psychologist, 43(3), 194–204. Kirkpatrick, H., Vogel, M. E., & Nyman, S. (2011). Changes psychologists must make to be successful in integrated primary care. Register Report, Fall. Klimoski, R., & Mohammed, S. (1994). Team mental model: Construct or metaphor? Journal of Management, 20(2), 403-437. Kniffin, K. M., & Wilson, D. S. (2010). Evolutionary perspectives on workplace gossip: Why and how gossip can serve groups. Group & Organization Management. Koudenburg, N., Postmes, T., & Gordijn, E. H. (2013). Conversational flow promotes solidarity. PloS one, 8(11), e78363. Laing, M. (1993). Gossip: Does it Play a Role in the Socialization of Nurses? The Journal of Nursing Scholarship, 25: 37– 44. Larkin, G. L., Moskop, J., Sanders, A., Derse, A. (1994). The emergency physician and patient confidentiality: A review. Annals of Emergency Medicine, 24(6), 1161–1167. Lashinsky, A. (2012). Inside Apple: How America's Most Admired--and Secretive--Company Really Works. Hachette UK.

82 Levine, J. M., & Moreland, R. L. (1999). Knowledge transmission in work groups: Helping newcomers to succeed. Shared cognition in organizations: The management of knowledge, 267- 296. Levine, M., Prosser, A., Evans, D., & Reicher, S. (2005). Identity and emergency intervention: How social group membership and inclusiveness of group boundaries shape helping behavior. Personality and Social Psychology Bulletin, 31(4), 443-453. Livingstone, A., & Haslam, S. A. (2008). The importance of social identity content in a setting of chronic social conflict: Understanding intergroup relations in Northern Ireland. British Journal of Social Psychology, 47(1), 1-21. Loi, R., Chan, K. W., & Lam, L. W. (2014). Leader–member exchange, organizational identification, and job satisfaction: A social identity perspective. Journal of Occupational and Organizational Psychology, 87(1), 42-61. Loi, R., Mao, Y., & Ngo, H. Y. (2009). Linking leader-member exchange and employee work outcomes: The mediating role of organizational social and economic exchange. Management and Organization Review, 5(3), 401-422. Lopez, S. J., & Prosser, E. (1999). Preparing psychologists: More focus on training psychologists for a future in evolving health- care delivery systems (Elliot & Klapow, 1997) (or, I’ll hire your graduates if...). Journal of Clinical Psychology in Medical Settings, 6(3), 295–301. Lyren, A., Kodish, E., Lazebnik, R., & O’Riordan, M. A. (2006). Understanding confidentiality: perspectives of African American adolescents and their parents. Journal of Adolescent Health, 39(2), 261-265. Mackie, D. M., Worth, L. T., & Asuncion, A. G. (1990). Processing of persuasive in-group messages. Journal of Personality and Social Psychology, 58(5), 812-822. Mael, F., & Ashforth, B. E. (1992). Alumni and their alma mater: A partial test of the reformulated model of organizational identification. Journal of Organizational Behavior, 13(2), 103-123. Magnus, J.R., & Viswesvaran, C. (2005). Whistle-blowing in organizations: An examination of correlates of whistle-blowing intentions, actions and retaliation. Journal of Business Ethics, 62(3), 277–297. Magnusson, J., Oakley, L., & Townsend, J. (2007). Parents’ views on confidentiality and health advice for adolescents in general practice. Primary Health Care Research and Development 8(14): 121–127. Markel, H. (2004). History of medicine: on the Hippocratic Oath. New Engl J Med, 350, 2026. McAndrew, F. T., & Milenkovic, M. A. (2002). Of tabloids and family secrets: The evolutionary psychology of gossip. Journal of Applied Social Psychology, 32(5), 1064-1082. McAndrew, F. T., Bell, E. K., & Garcia, C. M. (2007). Who do we tell and whom do we tell on?

83 Gossip as a strategy for status enhancement. Journal of Applied Social Psychology, 37(7), 1562- 1577. McGarty, C., Haslam, S. A., Hutchinson, K. J., & Turner, J. C. (1994). The effects of salient group memberships on persuasion. Small Group Research, 25(2), 267-293. McKinstry, B., Watson, P., Pinnock, H., Heaney, D., & Sheikh, A. (2009). Telephone consulting in primary care: a triangulated qualitative study of patients and providers. Br J Gen Pract, 59(563), e209-e218. Michelson, G., Van Iterson, A., & Waddington, K. (2010). Gossip in organizations: Contexts, consequences, and controversies. Group & Organization Management. Mlinek, E. J., & Pierce, J. (1997). Confidentiality and privacy breaches in a university hospital emergency department. Academic Emergency Medicine, 4(12), 1142-1146. Morton, T., Wright, R., Peters, K., Reynolds, K. J., & Haslam, S. A. (2012). Social identity and the dynamics of organizational communication. Handbook of intergroup communication, 319-330. Moyer, M., & Sullivan, J. (2008). Student risk-taking behaviors: When do school counselors break confidentiality? Professional School Counseling, 11(4), 236–245. Murphy, M. J., DeBernardo, C. R., & Shoemaker, W. E. (1998). Impact of managed care on independent practice and professional ethics: A survey of independent practitioners. Professional Psychology: Research and Practice, 29, 43–51. Nicholson, H. (2001). The new world on gossip: We all love to hear the new scoop on famous personalities and our friends. The latest dish is that back-fence conversations are essential to society. Psychology Today, New York, 34(3), 40-45. Nocera, J. (2008). Apple’s culture of secrecy. New York Times, 26. Oakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality. Blackwell Publishing. Olsen, J. C., & Sabin, B. R. (2003). Emergency department patient perceptions of privacy and confidentiality. The Journal of Emergency Medicine, 25(3), 329-333. Ormrod, J., & Ambrose, L. (1999). Public perceptions about confidentiality in mental health services. Journal of Mental Health, 8(4), 413–421. Paulsel, M. L., Richmond, V. P., McCroskey, J. C., & Cayanus, J. L. (2005). The relationships of perceived health professionals’ communication traits and credibility with perceived patient confidentiality. Communication Research Reports, 22(2), 129-142. Paulsen, N., Graham, P., Jones, E., Callan, V. J., & Gallois, C. (2005). Organizations as intergroup contexts: communication, discourse, and identification. Intergroup communication: Multiple perspectives, 165-188. Pendrak, R. F., & Ericson, R. P. (1998). Information technologies need to protect patient

84 confidentiality. Healthcare Financial Management, 52(10), 66-69. Pérez-Cárceles, M. D., Pereniguez, J. E., Osuna, E., & Luna, A. (2005). Balancing confidentiality and the information provided to families of patients in primary care. Journal of Medical Ethics, 31(9), 531-535. Petronio, S., Ellemers, N., Giles, H., & Gallois, C. (1998). (Mis) communicating across boundaries interpersonal and intergroup considerations. Communication Research, 25(6), 571-595. Pinta, E. R. (2009). Decisions to breach confidentiality when prisoners report violations of institutional rules. Journal of the American Academy of Psychiatry and the Law Online, 37(2), 150-154. Pope, K. S., & Vetter, V. A. (1992). Ethical dilemmas encountered by members of the American Psychological Association: a national survey. American Psychologist, 47(3), 397. Postmes, T. (2003). A social identity approach to communication in organizations. In S. A. Haslam, D. van Knippenberg, M. J. Platow, & N. Ellemers (Eds.), Social identity at work: Developing theory for organizational practice (pp. 81-97). Philadelphia, PA: Psychology Press. Postmes, T., & Branscombe, N. R. (Eds.). (2010). Rediscovering social identity. Psychology. Postmes, T., Haslam, S. A., & Swaab, R. I. (2005). Social influence in small groups: An interactive model of social identity formation. European Review of Social Psychology, 16(1), 1-42. Postmes, T., Tanis, M., & De Wit, B. (2001). Communication and commitment in organizations: A social identity approach. Group Processes & Intergroup Relations, 4(3), 227-246. Pouillot, P. (2011). Data Security, Strategic Risk Executive Report, Nov. 2011, 1-13. Pratt, M. G. (1998). Central questions in organizational identification. Identity in organizations, 171-207. Pratt, M. G. (2001). Social identity dynamics in modern organizations: An organizational psychology/organizational behavior perspective. Social Identity Processes in Organizational Contexts, 13-30. Prewitt, K. (2011). Why it matters to distinguish between privacy & confidentiality. Journal of Privacy and Confidentiality, 3(2), 3. Pritts, J. (2008). The importance and value of protecting the privacy of health information: roles of HIPAA privacy rule and the Common Rule in health research. Institute of Medicine. Rabinovich, A., & Morton, T. A. (2010). Who says we are bad people? The impact of criticism source and attributional content on responses to group-based criticism. Personality and Social Psychology Bulletin, 36(4), 524-536. Rae, W.A., Sullivan, B. S., Peña Razo, N., George, C. A., & Ramirez, E. (2002). Adolescent health risk behavior: When do pediatric psychologists break confidentiality? Journal of Pediatric Psychology, 27(6), 541–549.

85 Reicher, S., Spears, R., & Haslam, S. A. (2010). The social identity approach in social psychology. Sage identities handbook, 45-62. Riketta, M. (2005). Organizational identification: A meta-analysis. Journal of Vocational Behavior, 66(2), 358-384. Robinson, I. (1991). Confidentiality for whom? Social Science & Medicine, 32(3), 279-286. Saks, A. M., & Ashforth, B. E. (1997). Socialization tactics and newcomer information acquisition. International Journal of Selection and Assessment, 5, 48 – 61. Samet, J. H., Winter, M. R., Grant, L., & Hingson, R. (1997). Factors associated with HIV testing among sexually active adolescents: A Massachusetts survey. Pediatrics, 100(3), 371-377. Sankar, P., Moran, S., Merz, J. F., & Jones, N. L. (2003). Patient perspectives on medical confidentiality: A review of the literature. Journal of General Internal Medicine, 18(8), 659– 669. Shardlow, S. (1995). Confidentiality, accountability and the boundaries of client-worker relationships. Ethical issues in social work, 65-83. Sherif, C. W., & Sherif, M. (Eds.). (1967). Attitude, ego-involvement, and change. Wiley. Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-positive psychology undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22(11), 1359-1366. Singer, E., Von Thurn, D. R., & Miller, E. R. (1995). Confidentiality assurances and response a quantitative review of the experimental literature. Public Opinion Quarterly, 59(1), 66-77. Stapleton, A. B., Hankes, D. M., Hays, K. F., & Parham, W. D. (2010). Ethical dilemmas in sport psychology: A dialogue on the unique aspects impacting practice. Professional Psychology: Research and Practice, 41(2), 143a. Stiles, P. G., & Petrila, J. (2011). Research and confidentiality: Legal issues and risk management strategies. Psychology, Public Policy, and Law, 17(3), 333. Stürmer, S., & Snyder, M. (2009). Helping “Us” versus “Them”. Towards a group-level theory of helping an altruism across group boundaries. In S. Stürmer & M. Synder (Eds.), The psychology of pro-social behaviour: group processes, intergroup relations and helping (pp. 33- 58). Oxford, UK: Blackwell. Sullivan, J. R., Ramirez, E., Rae, W. A., Razo, N. R., & George, C. A. (2002). Factors contributing to breaking confidentiality with adolescent clients: A survey of pediatric psychologists. Professional Psychology: Research and Practice, 33(4), 396. Suzuki, S. (1998). In-group and out-group communication patterns in international organizations: Implications for social identity theory. Communication Research, 25(2), 154-182. Swaab, R. I., Postmes, T., & Spears, R. (2008). Identity formation in multiparty negotiations.

86 British Journal of Social Psychology, 47(1), 167-187. Swaab, R., Postmes, T., van Beest, I., & Spears, R. (2007). Shared cognition as a product of, and precursor to, shared identity in negotiations. Personality and Social Psychology Bulletin, 33(2), 187-199. Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology, 33(1), 1-39. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. The social psychology of intergroup relations, 33(47), 74. Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorization and intergroup behaviour. European Journal of Social Psychology, 1(2), 149-178. Taylor, D. M., Chappell-Lawrence, J., & Graham, I. S. (1996). Facsimile communication between emergency departments and GPs, and patient data confidentiality. The Medical Journal of Australia, 167(11-12), 575-578. Taylor, L., & Adelman, H. S. (1998). Confidentiality: Competing principles, inevitable dilemmas. Journal of Educational & Psychological Consultation, 9(3), 267 – 275. Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude-behavior relationship: A role for group identification. Personality and Social Psychology Bulletin, 22(8), 776-793. Trowell, J. (2001) ‘Confidentiality and Child Protection’, in C. Cordess (ed.) Confidentiality and Mental Health, pp. 85–94. London: Jessica Kingsley. Turner, J. C. (1999). Some current issues in research on social identity and self-categorization theories. Social identity: Context, commitment, content, 6-34. Turner, J. C., & Tajfel, H. (1986). The social identity theory of intergroup behavior. Psychology of Intergroup Relations, 7-24. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Oxford: Blackwell. Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarty, C. (1994). Self and collective: Cognition and social context. Personality and Social Psychology Bulletin, 20, 454-454. Turner, J.C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel (Ed.) Social identity and intergroup relations (pp. 15–40). Cambridge. UK: Cambridge University Press. Tyler, T. R., & Blader, S. L. (2000). Cooperation in groups: Procedural justice, social identity, and behavioral engagement. Psychology Press. van Dick, R., & Haslam, S. A. (2012). Stress and well-being in the workplace: Support for key propositions from the social identity approach. Van Knippenberg, D. (2000). Work motivation and performance: A social identity perspective. Applied Psychology, 49(3), 357-371.

87 van Leeuwen, E., & Täuber, S. (2010). The strategic side of outgroup helping. The psychology of prosocial behavior: Group processes, Intergroup relations, and Helping, 81-99. Van Liew, J. R. (2012). Balancing confidentiality and collaboration within multidisciplinary health care teams. Journal of Clinical Psychology in Medical Settings, 19(4), 411–417. Waddington, K. (2005). Using diaries to explore the characteristics of work-related gossip: Methodological considerations from exploratory multimethod research. Journal of Occupational and Organizational Psychology, 78, 221-236. Warren, S. D., & Brandeis, L. D. (1890). The right to privacy. Harvard law review, 193-220. Weick, K. E., & Roberts, K. H. (1993). Collective mind in organizations: Heedful interrelating on flight decks. Administrative Science Quarterly, 38(3), 357–381. Wert, S. R., & Salovey, P. (2004). A social comparison account of gossip. Review of General Psychology, 8(2), 122. Westin, A. F. (1968). Privacy and freedom. Washington and Lee Law Review, 25(1), 166. Whetten-Goldstein, K., Nguyen, T. Q., & Sugarman, J. (2001). So much for keeping secrets: the importance of considering patients' perspectives on maintaining confidentiality. AIDS care, 13(4), 457-465. Wilder, C. (1979). The Palo Alto group: Difficulties and directions of the interactional view for human communication research. Human Communication Research, 5(2), 171-186. Wilson, D. S., Wilczynski, C., Wells, A., & Weiser, L. (2000). Gossip and other aspects of language as group-level adaptations. The Evolution of Cognition, 347-365. Younggren, J. N., & Harris, E. A. (2008). Can you keep a secret? Confidentiality in psychotherapy. Journal of Clinical Psychology, 64(5), 589-600.