ABSTRACT Diane J. Majewski. THE ASSOCIATION BETWEEN DEANS OF STUDENTS’ ATTITUDES TOWARD SUICIDE AND THEIR BASES OF POWER (Under the direction of Dr. Michael Poock). Department of Educational Leadership, April 2010. This study investigated the associations between attitudes toward suicide and bases of power in university administrators known as Deans of Students (DOS). The National Association for Student Personnel Administrators, the professional association of higher education administrators in student affairs, provided the survey population for this study. Instrumentation included an electronic web-based survey comprising of 9 items from the Attitudes toward Suicide Questionnaire, 15 items from the Power Base Inventory, and 2 demographic questions pertaining to gender and highest degree earned. Results of the one-sample chi-square analysis found DOS’ expertise, legitimate, and reward powers is related their belief that suicide can be prevented. In addition, DOS’ expertise power is related to their belief that suicide should be condemned. Results of independent sample t-tests found no significance between DOS’ gender and attitudes toward suicide and gender and bases of power. THE ASSOCIATION BETWEEN DEANS OF STUDENTS’ ATTITUDES TOWARD SUICIDE AND THEIR BASES OF POWER A Dissertation Presented to the Faculty of the Department of Educational Leadership East Carolina University In Partial Fulfillment of the Requirements for the Degree Doctor of Education by Diane J. Majewski April 2010 THE ASSOCIATION BETWEEN DEANS OF STUDENTS’ ATTITUDES TOWARD SUICIDE AND THEIR BASES OF POWER by Diane J. Majewski APPROVED BY: DIRECTOR OF DISSERTATION:______________________________________ Michael Poock COMMITTEE MEMBER:_____________________________________________ Crystal Chambers COMMITTEE MEMBER:_____________________________________________ David Siegel COMMITTEE MEMBER:_____________________________________________ Lynn Roeder INTERIM CHAIR OF THE DEPARTMENT OF EDUCATIONAL LEADERSHIP: ______________________________________________ William Rouse, Jr. ACTING DEAN OF THE GRADUATE SCHOOL: ______________________________________________ Paul Gemperline ©Copyright 2010 Diane J. Majewski ACKNOWLEDGEMENTS It takes a village to complete a dissertation. Many individuals played a significant role in helping me complete my dissertation, and they deserve an infinite amount of gratitude. The individuals who went above and beyond the call of duty deserve acknowledgement. The efforts of my committee were instrumental in the completion of this research project. Michael Poock, my chair, thank you for your outstanding supervision and direction. I learned from you the value of being patient, but persistent, during the writing process. Crystal Chambers, David Siegel, and Lynn Roeder, I appreciate your astute attention to the endless details of this study. Throughout this experience, I relied on friends and colleagues for academic support. John Stiles, Jayne Geissler, and Maggie Olszewska thank you for reading drafts of various chapters, providing constructive feedback, and cheering me on! Gwen Joyner, thank you for attending to the formatting details I was too exhausted to face. Day in and day out, my office mates were there to applaud my successes and pick me up when I needed a boost. Nancy Ausherman, Traci Lynch, Sarah Raab, and Miranda Newman, thank you for looking after me. To my classmates and members of the Roundtable, I salute you! Laura Spivey, Karen Pearce, Morgan Daughety, and Dan Mayo your friendship remains close to my heart. Finally, my library buddy, Kendra Harris, it was our camaraderie that made this process pleasurable. Thank you for reminding me it is alright to be human. DEDICATION It has been said the best things in life come in threes. For that reason, I dedicate this research project to the three most influential women in my life, my mother, Marsha, my aunt, Mary, and my boss, Liz. You inspire me to always do my best. Thank you for believing in me! TABLE OF CONTENTS LIST OF TABLES………………………………………………………………. xiii CHAPTER 1: INTRODUCTION………………………………………………. 1 Statement of the Problem……………………………………………… 3 Conceptual Framework………………………………………………… 5 Research Question……………………………………………………... 7 Purpose of the Study…………………………………………………… 7 Significance of the Study………………………………………………. 8 Overview of Methodology……………………………………………… 8 Definitions……………………………………………………………….. 10 Limitations……………………………………………………………….. 12 Null Hypotheses………………………………………………………… 13 Conclusion………………………………………………………………. 14 CHAPTER 2: REVIEW OF LITERATURE…………………………………... 15 Suicide…………………………………………………………………… 15 Definitions of Suicide……………………………………………. 15 Psychological Theories of Suicide…………………………….. 18 Psychological understandings of suicide……………… 18 Sigmund Freud…………………………………………… 21 Karl Menninger…………………………………………… 22 Edwin Shneidman……………………………………….. 23 Sociological Theory of Suicide………………………………… 27 Social integration and social regulation……………….. 28 Four types of suicide…………………………………….. 29 Durkheim’s research on suicide………………………... 31 Religious Perspectives about Suicide………………………… 32 Christian and non-Christian perspectives…………….. 32 Religious systems……………………………………….. 34 Empirical data on suicide and religion………………… 36 Summary…………………………………………………………. 38 Suicide and College Students………………………………………… 39 Suicide Rates of College Students……………………………. 39 The Big Ten Student Suicide Study…………………………… 41 Suicidal Ideation versus Suicidal Attempts…………………… 43 Suicide ideation in college students………………….... 43 College students attempting suicide…………………… 45 Summary…………………………………………………………. 47 Deans of Students……………………………………………………… 48 Development of Deans of Students…………………………… 48 Dean of women…………………………………………... 49 Dean of men……………………………………………… 53 Overall plight of deans………………………………….. 56 One Dean of Students………………………………………….. 57 A Dean’s Perspective Post World War II……………………... 58 Authority of Deans of Students………………………………… 62 Deans of Students Today………………………………………. 65 Summary…………………………………………………………. 67 Conceptual Framework………………………………………………… 68 Individuals’ Attitudes Toward Suicide…………………………. 69 Measuring Attitudes Toward Suicide………………………….. 73 The suicide opinion questionnaire……………………... 73 Suicide attitude questionnaire………………………….. 76 Attitudes towards suicide questionnaire………………. 78 Social Power…………………………………………………….. 90 Advent of social power………………………………….. 91 Definition of social power……………………………….. 93 Definition of social influence……………………………. 96 The bases of social power……………………………… 98 Examining the social power bases…………………….. 101 Measuring the Bases of Power………………………………… 106 Original instruments used to measure power………… 106 Rahim leader power inventory………………………….. 107 Power base inventory…………………………………… 109 Summary…………………………………………………………. 112 Conclusion………………………………………………………………. 114 CHAPTER 3: METHODOLOGY……………………………………………… 115 Research Question…………………………………………………….. 115 Population……………………………………………………………….. 116 Instrumentation…………………………………………………………. 118 Attitudes Towards Suicide Questionnaire……………………. 119 Power Base Inventory…………………………………………... 120 Procedures and Implementation………………………………. 122 Procedure of obtaining research instruments………… 122 Internal review board procedures……………………… 124 Survey administration…………………………………… 124 Null Hypotheses………………………………………………………… 125 Data Analysis……………………………………………………………. 134 Conclusion………………………………………………………………. 135 CHAPTER 4: RESULTS………………………………………………………. 137 Introduction……………………………………………………………… 137 Participant’s Responses……………………………………………….. 137 Analysis of Data………………………………………………………… 150 Null Hypothesis One……………………………………………. 151 Null Hypothesis Two……………………………………………. 151 Null Hypothesis Three………………………………………….. 152 Null Hypothesis Four……………………………………………. 152 Null Hypothesis Five……………………………………………. 153 Null Hypothesis Six……………………………………………… 153 Null Hypothesis Seven…………………………………………. 154 Null Hypothesis Eight…………………………………………… 154 Null Hypothesis Nine……………………………………………. 155 Null Hypothesis Ten…………………………………………….. 156 Null Hypothesis Eleven…………………………………………. 156 Null Hypothesis Twelve………………………………………… 157 Null Hypothesis Thirteen……………………………………….. 157 Null Hypothesis Fourteen………………………………………. 158 Null Hypothesis Fifteen…………………………………………. 159 Null Hypothesis Sixteen………………………………………… 159 Null Hypothesis Seventeen…………………………………….. 160 Null Hypothesis Eighteen………………………………………. 160 Null Hypothesis Nineteen………………………………………. 161 Null Hypothesis Twenty………………………………………… 163 Summary of Significant Findings……………………………………… 164 Challenges with Instrumentation……………………………………… 165 Conclusion………………………………………………………………. 168 CHAPTER 5: DISUCSSION, IMPLICATIONS AND RECOMMENDATIONS………………………………………………………… 169 Introduction………………………………………………………………. 169 Participant Demographics……………………………………………… 169 Major Findings of the Study……………………………………………. 170 Discussion of the Results………………………………………………. 170 Finding 1: Expertise Power and Prevention of Suicide……… 171 Finding 2: Legitimate (Authority) Power and Prevention of Suicide……………………………………………………………. 172 Finding 3: Reward Power and Prevention of Suicide……….. 173 Finding 4: Expertise Power and Condemnation of Suicide… 174 Conceptual Framework………………………………………………… 176 Implications for Educational Leaders…………………………………. 178 Recommendations for Future Research……………………………… 181 Limitations of the Study………………………………………………… 183 Conclusion………………………………………………………………. 184 REFERENCES…………………………………………………………………. 186 APPENDIX A: ATTITUDES TOWARDS SUICIDE QUESTIONNAIRE (ATTS)…………………………………………………………………………... 205 APPENDIX B: POWER BASE INVENTORY (PBI)………………………… 215 APPENDIX C: NASPA MEMBERSHIP LIST REQUEST POLICY FOR RESEARCH PURPOSES…………………………………………………….. 218 APPENDIX D: NASPA MEMBERSHIP LIST REQUEST FORM…………. 219 APPENDIX E: CPP, INC. REQUEST FOR RESEARCH SUPPORT…….. 223 APPENDIX F: SIGNED CPP, INC. CONTRACT…………………………… 227 APPENDIX G: INSTITUTIONAL REVIEW BOARD PROCESSING FORM…………………………………………………………………………… 229 APPENDIX H: MODIFIED FORMAT OF ATTS AND PBI…………………. 230 APPENDIX I: MODIFIED SURVEY FORMAT OF ATTS AND PBI………. 231 LIST OF TABLES 1. Gender and Highest Degree Earned of Participants………………….. 139 2. Results of Chi-Square Tests for Hypotheses 1-18……………………. 166 CHAPTER 1: INTRODUCTION During the 2009 – 2010 academic year, the bodies of six Cornell University students were recovered from the bottom of the Ithaca campus gorges after they intentionally jumped off a bridge (Epstein, 2010). A south Florida college student blogging on the computer about his plan to commit suicide was found dead 12 hours later after taking an overdose of prescription medication (The Associated Press, 2008). Following multiple incidents of self-destructive behaviors on campus, a student at Massachusetts Institute of Technology consumed enough prescription medication to terminate her life (Scelfo, 2003). At New York University, three students committed suicide during a 40 day time frame by leaping from campus buildings (Scelfo). Suicide rates on college campuses have tripled since the 1950s (Brener, Hassan, & Barrios, 1999). Suicide ranks number two in the cause of death, after accidental death, among college students with approximately 1,100 students nationwide committing suicide annually (Appelbaum, 2006; Francis, 2003). Although student suicides represent less than one percent of the total undergraduate population of 17.5 million (U. S. Department of Education, National Center for Education Statistics, 2008), it is still a significant number of deaths. A single death on campus is viewed by the university community as a tragedy (Schwartz, 2006). Hence, Schwartz believes the public pays attention to suicide, and fully expects administrators at universities to examine the ways they attempt to identify students at risk for committing suicide. 2 Published studies on suicide and college students primarily focus on the demographic and psychological profiles of students who attempted suicide (Barrios, Everett, Simon, & Brener, 2000; Brener et al., 1999; Langhinrichsen- Rohling, Arata, Bowers, O’Brien, & Morgan, 2004; Silverman, Meyer, Sloane, Raffel, & Pratt, 1997; Westefeld, Homaifar, Spotts, Furr, Range, & Werth, 2005). The numbers of students who attempted suicide are well documented in the research (Konick & Gutierrez, 2005; Rudd, 1989; Silverman, 1993; Silverman et al.). The potential types and causes of suicidal thoughts in college students also received attention in the literature (Kisch, Leino, & Silverman, 2005; Westfeld et al., 2005). Additionally, the implications of psychological problems of hopelessness and depression, in relation to suicidal thoughts, continue to receive substantial attention (Leenaars, 2003). Deans of Students (DOS) are recognized as the campus authority responsible for the development of the whole student which includes being the first senior level administrator to respond to students who attempt suicide (Blue, 1972; Eddy, Chen, & Ball, 1988; Komives, Woodard, & Associates, 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). Power, including authority, is defined as the ability by which an individual, in this case DOS, convinces another person to comply with his or her wishes (French & Raven, 1968). University presidents grant substantial power to DOS who deal with matters concerning students in crisis (Appleton, Briggs, & Rhatigan, 1978; Komives et al., 2003; Nuss; Sandeen; Sandeen & Barr). In addition, DOS position of authority within 3 the institution’s organizational structure and their scope of responsibilities including monitoring the safety and well-being of students lend considerable power. These duties and authorities of DOS which have been investigated in literature have remained somewhat unchanged (Appleton et al., Komives et al., Nuss; Sandeen; Sandeen & Barr). As the campus authority responsible for the development of the whole student, DOS play a vital role in the methods used to identify students at risk for committing suicide (The Jed Foundation, 2006). Specifically, DOS oversee the programming and procedures that attempt to identify students at risk for suicide (Bost & Ballou, 1992; Francis, 2003; Grayson, 1994; The Jed Foundation; Kitzrow, 2003; Meilman, Pattis, & Kraus-Zeilmann, 1994; Silverman, 1993). Additionally, the responsibility of organizing a campus-wide social marketing program that promotes students’ sense of belonging and de-stigmatizes the use of mental health services rests with DOS (The Jed Foundation). Ultimately, monitoring the health of students is the responsibility of these student affairs administrators (Sandeen, 1991). Statement of the Problem The university environment is perceived to be an ideal setting for studying and preventing suicide (The Jed Foundation, 2006). Whitaker (1986) points out that the tight-knit structure and organization of the campus living environment frees university administrators to experiment with various methods that attempt to identify students at risk for committing suicide. In addition, most universities 4 provide easy access to quality medical and psychological services, the ideal place for monitoring students at risk for committing suicide (Whitaker). In the face of these observed potential benefits available to students in crisis, university administrators appear to find it challenging to identify students at risk for suicidal behaviors (Brener et al., 1999; Gutierrez, Osman, Kopper, Barrios, & Bagge, 2000; Kisch et al. 2005; Langhinrichsen-Rohling et al., 2004; Suicide Prevention Resource Center, 2004). Researchers have found that university administrators struggle with how to recognize students at risk for committing suicide which suggests they might have difficulties with creating a campus community that encourages students in crisis to seek help before they take their life (Cook, 2007; Westefeld et al., 2005). Overall, there appears to be a lack of consensus in the identification and application of standard practices for identifying students at risk for committing suicide (The Jed Foundation, 2006). The methods used to identify students at risk for committing suicide at colleges and universities in the United States appear to lack universally accepted standards of practice (The Jed Foundation, 2006; Suicide Prevention Resource Center, 2004). Rather, the attempts to identify students at risk for committing suicide rests with university administrators who hold the title, DOS, and as the literature states, there is inconsistency regarding the manner in which DOS respond to students who consider committing suicide (Bost & Ballou, 1992; Francis, 2003; Grayson, 1994; The Jed Foundation; Kitzrow, 2003; Silverman, 1993; Suicide Prevention Resource Center). 5 Likewise, as first responders to students in crisis, DOS are in a position of authority where they might be able to influence the methods used to identify students at risk for committing suicide (Blue, 1972; Eddy et al., 1988; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). It is unknown whether or not efforts to identify students at risk for committing suicide are guided by the DOS opinions about suicide and position of authority in their role as an administrator in student affairs. A better understanding of DOS’ perceived power and opinion about suicide is critical since it appears DOS are responsible for implementing the best methods used to identify students who might commit suicide. Conceptual Framework Serving as direct university advocate and care-taker of undergraduate students (Komives et al., 2003), DOS play a critical role in identifying students in crisis who might consider suicide as a solution to their problems (The Jed Foundation, 2006). Thus, the conceptual framework used for this study is twofold. First, the three factor model of attitudes toward suicide of acceptance, condemnation, and prevention is used to identify the Deans of Students’ attitudes toward suicide (Salander-Renberg, Hjelmeland, & Koposov, 2008). Second, the six base social power typology of coercive, expert, informational, legitimate, referent, and reward is used to identify the bases of power utilized by DOS (French &d Raven, 1968). This section provides a comprehensive description of this conceptual framework. 6 The study of attitudes toward suicide appears complex and unpredictable (Bayet, 1922; Gibbs, 1968; Platt, 1989; Sale, Williams, Clark, & Mills, 1975). The research acknowledges that attitudes toward suicide typically reflect theoretical perspectives including psychological theories, sociological theory, and religious perspectives (Holmes & Holmes, 2005; Leenaars, 2003; Maris, Berman, & Silverman, 2000). In their effort to measure attitudes toward suicide, Salander- Renberg et al. (2008) note that factors of acceptance, condemnation, and prevention include principles related to psychological, sociological, and religious perspectives of suicide. The complexities of human behavior, whether viewed from a psychological, sociological, or religious lens, make it difficult to predict or determine why an individual commits suicide (Leenaars, 2003). Next, the study of the bases of power (French & Raven, 1959, 1968) attempts to understand how individuals in a position of authority influence persons in their environment. Power is an individual’s application of authority over persons with the intent to sway the persons to conform to the wishes of the individual (Cartwright, 1959). French and Raven believe an individual can wield power through compensation (reward) or punishment (coercion). Power can be achieved through the individual’s position of authority (legitimate), and through the respect, exceptional knowledge, and admiration the persons have for the individual in charge (expert, informational, referent). As a student affairs administrator responsible for the development of the whole student, DOS are capable of asserting these various types of power when determining the means 7 by which student affairs professionals identify students at risk for committing suicide. Ascertaining the bases of power preferred by DOS, in conjunction with determining their attitudes toward suicide, might offer an understanding of the influence or authority DOS have on the manner in which they persuade professionals within the university community to identify students at risk for committing suicide. It is plausible that efforts made by individuals within the university community to identify students at risk for committing suicide are guided by DOS opinions about suicide, in part, because of DOS position of authority in their role as an administrator in student affairs. Research Question This research examined the association between French and Raven’s power base theory and the three factor model of attitudes toward suicide (Salander-Renberg et al., 2008) in student affairs administrators known as DOS. The overarching research question is: What is the association between attitudes toward suicide and bases of power in DOS? Purpose of the Study The purpose of this study is to understand the association between attitudes toward suicide and bases of power in DOS. The literature confirms that administrators in student affairs acquire a level of power due, in part, to their position or role within the organization, and their personal power inherent within themselves. It is unknown whether or not efforts to identify students who might 8 commit suicide are guided by DOS opinions about suicide and position of authority. Therefore, a better understanding of the collective perceptions of DOS nationwide toward suicide in association with their perceived bases of power appears essential to understanding why it appears university administrators struggle to identify students who commit suicide. Significance of the Study This study has several implications for educational leaders. Most importantly, results of this study could impact the manner in which student affairs administrators approach identifying students at risk for committing suicide. The DOS attitudes toward suicide are likely to recognize different feelings, cognitions, and opinions concerning the act of committing suicide (Diekstra & Kerkhof, 1998). The social power bases established by French and Raven (1959, 1968) identify the authority deans believe they have over faculty, staff, and students (Erchul, Raven, & Whichard, 2001). Findings of this study attempt to investigate the associations between attitudes toward suicide and bases of power held by DOS in their role as lead student affairs administrator responsible for the safety and well-being of students. Overview of Methodology The Attitudes toward Suicide Questionnaire (ATTS) and the Power Base Inventory (PBI) were modified and posted electronically into the on-line survey system known as Perseus for the purposes of investigating the research question. Research participants consisted of members from Student Affairs 9 Administrators in Higher Education (NASPA, n.d.) whose employment title is DOS. NASPA is the international professional organization for student affairs leaders in higher education (NASPA, n.d.). The Attitudes toward Suicide Questionnaire (ATTS), created by Salander- Renberg and Jacobsson (2003), is a self-administered, forced-choice questionnaire consisting of 61 items. The purpose of the ATTS is to identify the respondent’s attitudes and opinions about suicide (Salander-Renberg & Jacobsson, 2003). Nine statements from the ATTS were used to measure attitudes toward suicide. Attitude variables measured in this research include: (a) acceptance of suicide, (b) condemnation of suicide, and (c) prevention of suicide (Salander-Renberg & Jacobsson). The PBI, created by Thomas and Thomas (1991), is a self-administered, forced-choice inventory consisting of thirty items. The purpose of the PBI is to measure the respondent’s power base. Scores are converted to percentile ranks according to the six bases of power (Elliott, 2004). The first fifteen paired statements of the PBI were used to measure participants’ perceived authority. Power base variables measured in this research include: (a) reward, (b) coercion (discipline), (c) legitimate (authority), (d) referent (goodwill), (e) expert, and (f) informational (Thomas & Thomas). Finally, two demographic questions were included in the survey. The demographic questions asked respondents for information about their gender 10 and highest degree earned. This study utilized the demographic information for gender. Definitions The following operational terms were applied to this research study as the basis for understanding the context of this research. The terms and corresponding definitions relate to the components of this study. Attitudes toward suicide – beliefs about suicide as defined by Salander- Renberg et al. (2008). Acceptance – an opinion that suicide is an understandable option when the individual believes no other option exists. Condemnation – an opinion that suicide is a forbidden act Prevention – an opinion that suicide can be avoided with intervention from others. Bases of Power – the formal taxonomy of power established by French and Raven (1959, 1968) which includes six sources of influence. Coercive power – influencing agent threatens punishment if the target does not comply with the request. Expert power – influencing agent relies on superior knowledge in order to get the target to comply with the request. Informational power – Influencing agent relies on a greater understanding of the facts than that of the target in order to get the target to comply with the 11 request. Originally included with expert power due to its similarities, Raven convinced French to separate the two in 1965, making informational power the sixth power type. Legitimate power – influencing agent authority is a result of the influencing agent’s position in the organization. Referent power – the target complies with the influencing agent out of respect and admiration for the influencing agent. Reward power – influencing agent offers some form of compensation to the target that complies with the request. Deans of Students (DOS) - manager, mediator, leader, and educator responsible for the development of the whole student (Sandeen, 1991). Influence – the ability to control, persuade, or sway. Influencing agent – the individual doing the controlling, persuading or swaying. National Association of Student Personnel Administrators (NASPA) – lead professional organization made up of student affairs professionals for the purpose of advocating and disseminating information on high quality standards of practice (NASPA, n.d.). Power – “…the induction of psychological forces by one entity b upon another a, and to the resistance to this induction set up by a” (Cartwright, 1959, p. 188). Power Base Inventory – survey instrument created by Thomas and 12 Thomas (1991) that measures the respondent’s power base. Social Power – a form of influence on an individual or an organization (Cartwright, 1965). Student Affairs Administrator – higher education professional charged with the social and emotional development of the whole student in support of the academic mission of the institution (Sandeen & Barr, 2006). Suicidal behavior – actions that suggest an individual might end his or her life. Suicidal ideation – thoughts of serving as the agent of one’s own death (American Psychological Association [APA], 2003). Suicide – Latin meaning “to kill oneself” (Fuse, 1997; Holmes & Holmes, 2005). Suicide Attitude Questionnaire (SUIATT) – survey instrument created by Diekstra and Kerkhof (1998) that identifies the respondent’s opinion about suicide. Target – the recipient of the influencing agent’s requests. Limitations Several limitations exist in this study. The representative sample completing the modified versions of the ATTS and the PBI consists of student affairs professionals with memberships to NASPA. Membership into NASPA is voluntary on the part of the DOS; therefore, data extracted from this study applies only to the individuals who participated in this study. In addition, the true 13 number of DOS registered with NASPA was difficult to obtain. According to K. Kruger (personal communication, October 9, 2009), various upper-level student affairs administrators might perform a multitude of duties usually performed by traditional DOS without possessing the working title, DOS. Therefore, the employment title, DOS, appears to be inconsistent in its application across institutions. Self-evaluation survey instruments like the ATTS and the PBI have limitations. The inability to measure whether or not the participants’ attitude reflects the individual’s actual behavior is a shortcoming of these instruments. Participants might respond in a socially desirable fashion rather than report their own perceptions. As a result, findings might be an underestimation of lenient attitudes toward suicide (Conner, Duberstein, Conwell, & Seidlitz, 2001). Finally, accurate data on the death of a student due to suicide is difficult to obtain (Kisch et al., 2005; Konick & Gutierrez, 2005; Silverman, 1993; Silverman et al., 1997; Westefeld et al., 2005). Consequently, the true number of students who die by suicide is unknown. Participants’ responses on the ATTS might be impacted by the knowledge that data on student suicide is inconsistent. Null Hypotheses This study identified DOS attitudes toward suicide in relationship to the social power bases used by the DOS. Null hypotheses (1 – 6) state there is no statistically significant association between the social power bases (reward, coercion, legitimate, referent, expert, and information) and the suicide attitude of 14 acceptance in DOS. Null hypotheses (7 – 12) state there is no statistically significant association between the social power bases (reward, coercion, legitimate, referent, expert, and information) and the suicide attitude of condemnation in DOS. Null hypotheses (13 – 18) state there is no statistically significant association between the social power bases (reward, coercion, legitimate, referent, expert, and information) and the suicide attitude of preventability in DOS. Null hypothesis 19 states there is no statistically significant difference between age and attitude toward suicide in DOS. Null hypothesis 20 states that there is no statistically significant difference between age and social power base in DOS. Conclusion The death of a student on campus is considered a traumatic event for the entire university community. Research suggests a long history of limited efforts by administrators at colleges and universities to successfully identify students at risk for committing suicide (The Jed Foundation, 2006; Meilman et al., 1994; Webb, 1986; Westefeld & Pattillo, 1987). Student development and student crisis management on university campuses are judged the responsibility of DOS (Blue, 1972; Eddy et al., 1988; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). This study attempts to add to the research by understanding DOS perceived attitudes toward suicide and positions of power as chief student affairs administrator responsible for the safety and well-being of students. CHAPTER 2: REVIEW OF THE LITERATURE This review of the literature encompasses four themes. First, a comprehensive look at suicide explores the definitions, theories, and perspectives of suicide. Second, evidence of research on suicide and college students investigates the incidence of suicide, suicide ideation, and suicidal attempts in college students. Third, as student affairs professionals, DOS serve as the primary caretaker of students responsible for identifying students who might commit suicide. A review of the roles, responsibilities, and authority of DOS proves essential. Fourth, a closer look at the conceptual framework of this study analyzes the exploration of individuals’ attitudes toward suicide, the measurement of attitudes toward suicide, the exploration of the social power bases, and the measurement of the social power bases. Suicide The study of suicide investigates why individuals take their own lives (Maris et al., 2000). Theories and perspectives on suicide attempt to grasp the meanings behind why individuals commit suicide (Maris et al., 2000). This section of the literature review is sorted into four sections: (a) definitions of suicide, (b) psychological theories of suicide, (c) sociological theory of suicide, and (d) religious perspectives on suicide. Definitions of Suicide Kisch et al. (2005) identified the three causes of death for young people aged 15-24 as unintentional injury, homicide, and suicide. The individual is no 16 longer alive; consequently, it is difficult to know whether the person died accidentally, by another’s hand, or by self injury. Barrios et al. (2000) found that young people involved with unintentional injury, homicide, or suicide exhibited suicidal thoughts and behaviors prior to dying. In an attempt to understand and distinguish suicide from the two other leading causes of death, Holmes and Holmes (2005), Freud (1917/1961), Menninger (1938), Shneidman (1985), and Durkheim (1951), and professionals from the medical and legal fields offer various definitions of suicide. The Latin root definition of suicide means “to kill oneself” (Fuse, 1997; Holmes & Holmes, 2005). Holmes and Holmes describe suicide as an “intentional act that is sometimes accompanied by gross circumstantial events that legitimize the action” (p. 15). Generally connected to a personal problem that appears inescapable, individuals might choose suicide as the solution to escaping the problems of life. Others report committing suicide for a social, political, or religious cause because the persons feel commanded to die for a cause. In either case, suicidal individuals report difficulties in dealing with enormous stress or conflict and suicide is viewed to be the only solution (Holmes & Holmes). The theories of suicide studied by suicidologists also define suicide. For example, Freud (1917/1961) described suicide as an aggressive act generated by the individual’s feelings of depression and anger. Karl Menninger (1938) defined suicide as a murder of the self, a wish to be killed, or a wish to die. Edwin 17 Shneidman (1985) described suicide as “a conscious act of self-induced annihilation, best understood as a multidimensional malaise in a needful individual who defines an issue for which suicide is perceived as the best solution” (Maris et al., 2000, p. 30). Emile Durkheim (1951) defined suicide as “applied to all cases of death resulting directly or indirectly from a positive or negative act of victim himself, which he knows will produce this result” (Maris et al., 2000, p. 30). Finally, the medical and legal fields use four points to define suicide (Maris et al., 2000). First, death by suicide requires a legal and medical document like a death certificate. Second, death by suicide is anticipated, usually not unintentional or accidental. Third, an individual commits suicide to oneself; a person killing a suicidal individual is viewed as an act of murder even if the suicidal individual wanted to die. Fourth, indirect or passive behaviors like failing to take medicine intended to save the individual’s life might result in suicide (Maris et al., 2000). Collectively, definitions of suicide discussed here share several common themes. Suicide is an act an individual commits to oneself not others. The act of suicide appears intentional and the individual knows the probable outcome is death. Moreover, a suicidal individual thinks suicide is the best solution to life’s problems. Emerging from the definitions of suicide, researchers have spent considerable effort on studying the possible psychological theories behind the act of ending one’s life. 18 Psychological Theories of Suicide The study of suicide from a psychodynamic view offers a myriad of theories which explore why an individual might choose to end his or her life. A summary of the eight psychological theoretical understandings of suicide by Leenaars (2003) is presented. Following is a discussion of theories of suicide from suicidologists, Sigmund Freud (1917/1961), Karl Menninger (1938), and Edwin Shneidman (1985). In addition, data on the number of people who committed suicide, as the result of a psychological stressor, is provided. Psychological understandings of suicide. A series of empirical studies by Leenaars (1988, 1989a, 1989b, 1996) produced a summary of eight psychological understandings of suicide. These studies analyzed the core aspects of suicide from the research of ten prominent suicidologists: Adler, Binswanger, Freud, Jung, Menninger, Kelly, Murray, Shneidman, Sullivan, and Zilboorg. The analyses yielded thirty-five psychological propositions for understanding suicide from which Leenaars reduced to eight. Leenaars published this list of eight psychological understandings in 2003. He described the first five concepts as components of the intra-psychic or personality and the last three as interpersonal or contextual features of the mind. The following is a summary of the eight psychological understandings of suicide as deduced by Leenaars (2003) and interpreted by other researchers. Unbearable Psychological Pain is a motivation for suicide when the emotional hurt is at a heightened state of agitation (Shneidman, 1985, 1993). The 19 individual feels rejected and helpless and thus desires to escape this pain (Menninger, 1938). The act of suicide functions as a relief from this unbearable pain (Murray, 1967). Cognitive Constriction describes the thought process of individuals who commit suicide as mentally narrow in focus and concrete (Shneidman, 1985). Right before death the individual appears to only think about the trauma that led to the decision to end life. It is this constricted behavior that enhances the danger of suicide (Shneidman, 1985). Indirect Expressions portray suicidal individuals as ambivalent (Freud, 1917/1961; Leenaars, 1988, 1993). An individual’s conscious awareness of the suicidal mind is small; therefore, the individual’s complicated and contradictory feelings about suicide tend to lead the individual toward death instead of life (Freud, 1917/1961; Leenaars, 1988, 1993). Inability to Adjust is the suicidal individual’s view of themselves as fragile and unable to change (Leenaars, 1988; Sullivan, 1962, 1964). Due to their powerlessness, they report feeling unable to cope with life’s stressors. This inability to adjust often results in suicidal individuals being diagnosed with a depressive disorder, a manic-depressive disorder, or an anxiety disorder (Leenaars, 1988; Sullivan, 1962, 1964). Ego is a person’s sense of individuality and is a key factor in the suicide formula (Murray, 1938). The strength of the ego has the potential to protect against suicide. However, when weakened by a long history of trauma like abuse 20 or failure, the ego is a contributing factor to the individual’s decision to commit suicide (Zilboorg, 1936). Interpersonal Relations are the significant problems suicidal individuals have in maintaining relationships with others (Murray, 1967). Suicide is linked to an unsatisfied or frustrated attachment need to others. The individual presents feelings of worry that remain present constantly (Murray, 1967). Rejection-Aggression is the belief that “Loss is central to suicide; it is, in fact, often a rejection that leads to pain and self-directed aggression. Suicide maybe veiled aggression—it may be murder in the 180th degree” (Leenaars, 2003, p. 9). Identification-Egression is when suicidal individuals experience an extreme attachment to a rejecting person (Freud, 1961) or lose social/emotional value like their youth or freedom (Zilboorg, 1936). The emotional attachment failed the suicidal individual, and he or she suffers an intense discomfort and yearns to escape (Freud, 1961; Zilboorg). Leenaars’ (2003) eight psychological theoretical understandings of suicide attempt to offer a psychological explanation of why an individual chooses to take his or her life. However, Leenaars (1999) believed these theoretical understandings of suicide are interactive and dynamic, but also uncertain. The complexities of human behavior make it difficult to predict or determine why an individual commits suicide (Leenaars, 2003). 21 Of the ten suicidologists mentioned in the Leenaars (2003) study, Sigmund Freud, Karl Menninger, and Edwin Shneidman were considered the lead researchers to publish theories on suicide (Maris et al., 2000). Their efforts to analyze suicide have become the foundation from which much of the literature on suicidology is based today. A brief overview of suicide theories by Freud (1917/1961), Menninger (1938), and Shneidman (1985) will provide an understanding of the psychological analysis of suicide. Sigmund Freud. The psychological study of suicide can be traced back to the work of Sigmund Freud (1917/1961). Much of his work stemmed from the psychodynamic viewpoint that linked suicide to an individual’s depressed state of mind and feelings of unhappiness combined with narcissistic behaviors associated with the act of suicide itself. Freud believed suicide was an aggressive act, and any attempt to understand suicide from a psychodynamic point-of-view required focusing on the role of aggression in the psyche. For instance, Freud thought hostility directed toward an external object, like a parent, was the cause of suicide (Freud, 1917/1961). Freud’s theory on the role of aggression in suicide laid the foundation for the psychodynamic views of suicide present today (Lees & Stimpson, 2008). His initial theory postulated that aggression was a by-product of the drive to sexual mastery; however, years later he modified his position stating that aggression was actually part of the self preservative instinct (Freud, 1915). Building on these theories, Freud concluded that suicide was the result of re-routed aggression 22 (Freud, 1917/1961). Aggression “represents the ego’s original reaction to objects in the external world” (Freud, 1917/1961, p. 252). People end their lives because they are devastated by the presence of this external object. Freud summarized by saying that individuals commit suicide only if they desire to kill another. Building on Freud’s psychodynamic method of suicide, he believed the three fundamental dynamic mechanisms of suicide complemented his initial findings regarding a suicidal individual’s aggression towards others (Lees & Stimpson, 2008). Thus, in keeping with Freud (1917/1961), suicide required aggression plus the three fundamental dynamic mechanisms of suicide: internalization, detachment from self, and acting out. First, the internalization of the individual’s aggression begins to weigh heavily on the individual’s psyche. Second, there is a change in the individual’s relationship with self. The individual’s ego transforms itself into an object that is disconnected from its true self. Third, suicide is a result of an individual acting out in the present in order to get relief from a childhood trauma (Lees & Stimpson). Freud continued to use these ideas associated with aggression and suicide to corroborate the importance of the role of aggression in suicide. Karl Menninger. Known to be Freud’s American counterpart, Karl Menninger (1938) also studied the psychological theories of suicide. Menninger believed people who committed suicide were “depressed, hopeless, and cognitively constricted” (Maris, Berman, Maltsberger, & Yufit, 1992, p. 71). Suicide, as interpreted by Menninger, occurred when an individual experienced 23 the loss of an important love object, like a spouse, who the individual internalized as part of his/her ego. This internalized behavior usually manifested into a form of depression. Menninger (1938) classified suicide into three fundamental dimensions: hate, depression, and guilt. Furthermore, Menninger believed suicide takes on one of three forms: (a) revenge, a wish to kill; (b) depression/hopelessness, a wish to die; and (c) guilt, a wish to be killed. More often than not individuals with internalized feelings of the ego for a loved one often felt intense depression following the death of the loved one. This level of depression included feelings associated with anger, rage, revenge, or a wish to kill. Hence, Menninger viewed suicide as “murder in the 180th degree” (Maris et al., 1992). Litman (1967) found support for Menninger’s forms of suicide when he determined that an individual’s wish to kill divides the individual’s ego resulting in feelings of regression. The suicidal individual harbors guilty feelings for wanting to murder a loved one, consequently experiencing the “wish to kill” and the “wish to be killed”. The suicidal individual’s ego is consumed with self hatred, the guilt becomes unbearable, and the individual wishes to die (Litman). Edwin Shneidman. Labeled the “founding father of suicidology,” Edwin Shneidman (1985) concluded that suicide was a reaction to the individual’s intolerable mental pain. He coined this intolerable pain “psychache” or extreme psychological anguish. The threat of suicide is high when the individual feels 24 “psychache (pain), press (stress), and perturbation (agitation)” when experienced at maximum intensity (Maris et al., 2000, p. 50). Unlike Freud (1917, 1961) and Menninger (1938), Shneidman (1985, 1993) did not connect suicide to the individual’s loss or anger with an external object that was tied to the individual’s ego. Instead, Shneidman concluded that suicide was a conscious, self-induced act initiated by an individual experiencing intense dissatisfaction with life, and the individual thought suicide was the only recognizable solution. Outcomes of his research on suicide produced the 10 commonalities of suicide or as Shneidman (1985, 1993) states, “the common features in human self-destruction” (Maris et al., 2000, p. 48). According to Shneidman (1985, 1993), the 10 commonalities of suicide are: 1. To seek a solution: Suicide from this perspective is a means to an end and the best way to solve a problem. Life is excruciating and death is considered the best solution (Maris, 1981). 2. Cessation of consciousness: Suicidal individuals are convinced that ending their lives will stop the mental anguish they feel. To relieve this mental pain suicidal individuals might turn to sleep or drug and alcohol intoxication in order to relieve their consciousness from the burden of constant anguish. 3. Intolerable psychological pain: Shneidman (1993) identified intolerable psychological pain as the key commonality in all suicides. The central 25 component of suicide is the individual’s feelings of psychological hurt. 4. Frustrated psychological needs: Suicide is likely to occur when one or more of the individual’s psychological needs are not being met. Individuals lacking psychological needs related to achievement and autonomy are prone to suicide, or individuals lacking psychological needs related to security and self-esteem might exhibit suicidal tendencies. 5. Hopelessness-helplessness: Individuals contemplating suicide are usually depressed; however, it is the feeling of hopelessness that most contributes to self-destructive behaviors. Suicidal people believe their quality of life will probably never improve so living is no longer an option. 6. Ambivalence: Individuals thinking about suicide are generally undecided about whether or not to end their lives. It is the individual’s destructive reaction to a crisis situation that tips the scale in favor of suicide. 7. Constriction: From a psychological perspective suicidal individuals tend to have a narrow view of the alternatives to suicide. 8. Egression: Individuals thinking about suicide view suicide as one of the best ways to end their problems. Other strategies attempted like the over-consumption of alcohol and drugs have failed, so they egress, or exit from life itself. 26 9. Communication of intent: Individuals thinking about suicide usually talk about the desire to kill themselves. However, most communications of intent result in false positives—these individuals frequently do not kill themselves. 10. Lifelong coping patterns: Individuals prone to suicide as a rule persistently engage in self-destructive behaviors. This destructive behavior tends to manifest a constant state of crisis for the individual which leads to a life long coping pattern of self-destruction. Shneidman (1985, 1993) and others view suicide as a deliberate act to end one’s life. The pain, stress, and agitation of life are incredibly intense, and the individual lacks coping skills to deal with the disappointments of life. As a result, suicide is considered the best option to end the pain. Several interesting findings are present. Despite their theoretical differences in the psychological theories of suicide, Freud (1917/1961), Menninger (1938), and Shneidman (1985) share several commonalities regarding their theoretical perspective in understanding suicide. For instance, they seemed to agree that individuals most at risk for suicide appeared to be detached from their psychological self, they experienced intense bouts of depression and hopelessness, and they report feeling as if they have lost at life. According to this psychodynamic perspective, suicide appears to be a solution to a deep rooted psychological pain (Freud, 1917/1961; Menninger, 1938; Shneidman, 1985). 27 The most eminent suicidologists report that the general premise behind the psychological theories of suicide is that individuals contemplating suicide do so because they believe suicide is the only solution to their severe psychological pain (Leenaars, 2003). These individuals appear to become detached from their emotional selves, react to this pain in an aggressive manner, become absorbed by feelings of hopelessness, and suffer from a deep seeded sense of loss (Leenaars, 2003). The psychological understandings of suicide are an attempt to recognize and explain why an individual might choose to end his or her life. Sociological Theory of Suicide Psychological theorists attempt to understand suicide from an emotional or behavioral perspective; sociological theorists believe suicide is linked to an individual’s experiences or interactions with society. In 1897, the work Le Suicide, by Emile Durkheim, was considered, and is still thought to be, one of the most decisive sociological perspectives on suicide (Adair, 2008; Berk, 2006; Gane, 2005; Graeff & Mehlkop, 2007). Durkheim (1951) believed suicide was the result of the level of control society had over an individual. Unlike Freud (1917/1961), Menninger (1938), or Shneidman (1985), Durkheim believed psychological factors appeared to have no impact on suicide. This section provides a comprehensive review of: (a) the premise of Durkheim’s theory of suicide, (b) the purpose of social integration and social regulation, (c) the four types of suicide, and (d) a review of research conducted 28 by Durkheim. In addition, Fuse (1997) and Berk offer their interpretation of Durkheim’s findings throughout this discussion. Social integration and social regulation. The premise behind Durkheim’s sociological theory of suicide is the belief that suicide arose from societal pressures and influence (Durkheim, 1951; Fuse, 1997; Holmes & Holmes, 2005). Durkheim labeled these enormous stressors as internal and external personality stressors responsible for the individual’s decision to commit suicide (Holmes & Holmes). Hence, suicide is a direct result of social factors like the attachments the individual has to society and social groups, not psychological factors that contribute to insanity or alcoholism (Fuse). Credited as the first sociologist to construct a comprehensive theory of suicide, Durkheim attempted to identify the differences in suicide rates between various social groups (Berk, 2006). Durkheim believed each person functioned within society in two dimensions: integration—the manner in which the individual binds to the values of the social group, and regulation—the level at which the individual tolerates the laws of society (Durkheim, 1951; Holmes & Holmes, 2005; Lester). The social structures of integration and regulation might explain how well an individual fits into society and accepts the conditions of the individual’s society (Durkheim; Holmes & Holmes). Individuals feeling isolated from society or individuals refusing to accept society’s standards react to this problem by harming themselves (Adair, 2008). Behaviors that are associated 29 with integration and regulation that are too strong or too weak could mean the individual is prone to suicide (Durkheim; Holmes & Holmes). Four types of suicide. The various levels of social integration and social regulation yielded four possible types of suicide: (a) altruistic suicide, (b) anomic suicide, (c) fatalistic suicide and (d) egoistic suicide (Adair, 2008; Durkheim, 1951; Fuse, 1997; Holmes & Holmes, 2005). Altruistic suicidal experiences are described as being highly integrated with their social group and highly regulated with their society. Integration and regulation occur at a level where the individual loses his sense of identity. These people are willing to sacrifice their lives for the cause of the group (Adair; Durkheim; Fuse; Holmes & Holmes). Suicide bombers are an example of altruistic suicide (Holmes & Holmes). High social integration with their social group is a characteristic anomic suicide shares with altruistic suicide. However, it has been found individuals believed to have an anomic suicidal experience report a low level of social regulation as opposed to high levels of social regulation found in the altruistic counterpart (Adair, 2008; Durkheim, 1951; Fuse, 1997; Holmes & Holmes, 2005). As a result, this low social regulation, as described by Durkheim, is a disturbance in the individual’s social equilibrium. When serious adjustments to the environment or unexpected changes to the environment occur, the individual is likely to self destruct (Fuse). Anomic suicide is the direct result of extreme social pressures that are so intense, the only way out, according to the individual, is to end life (Holmes & Holmes). Two extreme examples of anomic suicide are the 30 intentional death of a professional athlete who is unable to handle the financial and social pressures of wealth, or the intentional death of an individual who lost his or her wealth during the Great Depression. These people saw no way out and committed suicide (Holmes & Holmes). Fatalistic suicide is the opposite of anomic suicide. Fatalistic suicides involve individuals who experience very low levels of social integration but very high levels of social regulation (Adair, 2008; Durkheim, 1951; Holmes & Holmes, 2005). Disconnected from the population and hyper-regulated by their environment, these people report feeling unrewarded by life and give up on living (Holmes & Holmes). A terminally ill individual or an individual being abused physically and emotionally by a partner are two examples of potential fatalistic suicide situations (Holmes & Holmes). Lastly, low social integration and low social regulation sometimes result in an egoistic suicide (Adair, 2008; Durkheim, 1951; Fuse, 1997; Holmes & Holmes, 2005). The polar reverse of altruistic suicide, egoistic suicidal individuals are detached from the social order and fail to embrace the values of society. The center of civilization is the individual, thus no satisfaction is gained from the achievements of society (Holmes & Holmes). Failure is the individual’s responsibility and is not a reflection of the populace (Fuse). These individuals are self centered rather than group centered, leaving the individual to bear the burden of failure alone (Adair; Durkheim; Fuse; Holmes & Holmes). Interestingly, identifying the contrast between egoistic suicide and anomic suicide, Fuse 31 stated, “The egoistic suicide may find life unbearable because of excessive self- discipline; the anomic suicide may find life unbearable because of inadequate self-discipline” (p. 88). Durkheim and Fuse believed academic scholars have a high probability of succumbing to egoistic suicide. Durkheim’s research on suicide. Research conducted by Durkheim yielded several interesting observations. Demographic results indicated that suicide was more prevalent among men than women; among Protestants than Jews or Catholics; among widowed, divorced, or single individuals than married individuals; among highly skilled professionals than laborers; and among military personnel than civilians (Fuse, 1997). Suicide rates usually decreased when a nation was at war but increased during times of economic prosperity (Fuse). According to Durkheim, suicide was considered a social phenomenon, and individuals prone to suicide did so because of circumstances in their society, not because they were predisposed to suicide (Fuse). These observations reported by Durkheim provided other researchers the chance to interpret his findings. Fuse (1997) found that throughout Durkheim’s research he was unable to determine or predict why certain social groups were more prone to suicide than others. He could only postulate that a person’s reason for suicide varied because the level at which the person was regulated by or integrated into group life appeared to impact the person’s motivation to consider suicide. The decision to commit suicide was, therefore, dependent on the level at which the individual was integrated into society (Fuse). 32 Overall, Durkheim offered a theory of suicide based on the social surroundings and conditions experienced by the individual (Fuse, 1997). Individuals disconnected from or disappointed with their social group appeared most at risk for harming themselves (Durkheim, 1951). Despite the evidence presented by Fuse and Berk (2006) which indicated that Durkheim’s theory failed to predict suicide, Adair (2008), Durkheim, and Holmes and Holmes (2005) found that an individual’s level of regulation and integration into society was a contributing factor to the individual’s decision to commit suicide. Religious Perspectives about Suicide Psychological and sociological theorists study suicide from behavioral or societal perspectives. Nonetheless, the study of suicide and religion has also received substantial attention in the literature. The religious perspective contends, “Suicide is the deliberate choice and successful effort to end one’s life, regardless of motives, circumstances, or methods used” (Clemons, 1990, p. 44). A brief review of the various theological perspectives about suicide including: (a) the Christian and non-Christian perspectives about suicide, (b) the religious systems that effect suicide, and (c) the empirical data on religion and suicide is discussed. Christian and non-Christian perspectives. There are numerous perspectives about suicide in the non-Christian and Christian communities. For example, Christians, Jews, and Muslims trust the acceptance of their creator’s plan, which includes suffering on earth because individuals faithful to their beliefs 33 are rewarded with spiritual maturity for following their intended path (Maris et al., 2000). On the other hand, Hindus believe individuals are “reborn”, and Buddhists think the body does not reflect the contents of the true soul. Finally, religious cults like Jonestown and Heaven’s Gate view suicide as an honorable right of passage to the next level of living. The emphasis in the literature pays particular attention to the perspectives of suicide in the Christian, Jewish, and religious cult’s faiths (Maris et al., 2000). Within the Christian and Jewish faiths to undertake suicide is considered an offensive act toward the plan of the higher authority being worshiped. An individual who ends his/her life is viewed as selfish in religious contexts (Maris et al., 2000). Suicide was affirmed a sinful act, as interpreted in the Bible, when St. Augustine’s City of God (412 AD) and the Christian Council of Braga (563 AD) confirmed that the sixth commandment, in the Bible, “thou shall not kill”, included suicide and homicide (Maris et al., 2000; McLaughlin, 2007). Other biblical references have offered interpretations of suicide as a sinful act. For instance, the book of Deuteronomy states that God has the power to create life and kill life; humans do not possess this power (Clemons, 1990). The host of biblical references to suicide as a sinful act has the potential to impact an individual’s perspective of suicide (Clemons). However, death by suicide was not always viewed as selfish and offensive. The Old Testament of the Christian Bible did not initially forbid suicide (McLaughlin, 2007). In fact, the Christian Bible recounted a handful of suicides 34 without censure (Maris et al., 2000). For example, the Bible called for suicide in situations that involved military defeat, shame or justice, and as a political solution to settle tribal disputes. Present theologians debate whether or not Jesus was a “sacrificial suicide” (Maris et al., 2000). The premise that suicide from the religious perspective is viewed as a vile act against the being the individual worships appears to be consistent among many theological denominations (Maris et al., 2000). Life, not death, seems to be more revered and celebrated by individuals practicing a particular faith. Along with the religious perspectives of suicide, religion appears to influence the rate of suicide within religious communities (Stack, 2001). Religious systems. According to Stack (2001), there are three religious systems that might affect suicide. First, religious integration has been demonstrated to lower the risk of suicide. Religious integration is the practice of individuals accepting their inferior role to the religious beliefs and practices of their faith. By taking on this subordinate role, individuals are protected from experiencing egotistic behaviors that tend to cause suicidal behavior. For example, research by Durkheim found that Catholics overall had lower suicide rates than Protestants. He attributed this lower suicide rate to the Catholic’s subordinate behaviors that provide order to life, reduce stress, and reduce the likelihood of suicide (Stack, 2001). Second, religious commitment or a promise to obey the beliefs of the religion tends to decrease the risk of suicide (Stack, 2001). Stack (1983) and 35 Stark, Doylen, and Rushing (1983) devised the religious commitment theory in response to these beliefs. They maintain that few core religious practices might lower suicide. For example, divorce and unemployment might be more bearable with the understanding that perseverance during these difficult circumstances will yield eternal salvation. Despite what might be perceived as sinful behavior according to the religion, commitment to the faith generally results in salvation (Stack, 1983; Stark et al., 1983). Studies by Stack, and Stark et al. confirmed that populations committed to their religious communities saw lower suicide rates than those with no religious commitment. Likewise, populations with no religious affiliations saw an increase in the rate of suicide. In a study by Hasselback, Lee, Yang, Nichol, and Wigle (1991), the Canadian census data determined that increase in the numbers of people reporting no religious affiliation was associated with the increase in the number of suicides by 3.2%. Third, religious networks or collective support groups with others that practice a similar faith are apt to observe lower suicide rates (Stack, 2001). A social support network among church members safeguards them from life’s crises. For example, countries like Egypt, the Philippines, Mexico and Ireland report low suicide rates because of their strong social support network in the Muslim and Catholic faiths (Maris et al., 2000). Furthermore, Pescosolido and Georgianna (1989) reviewed the suicide data for approximately 400 counties in the southern part of the United States and found that the greater the concentration of practicing Evangelical Baptists, the lower the suicide rates for 36 those counties. Stack believes that low suicide risk is directly proportionate to the strength of friendships experienced by the individuals in the church. Empirical data on suicide and religion. The empirical data on suicide and religion is meager. What little data exist seem to suggest that religion tends to protect individuals from suicide, but it is not a clear causal connection, other variables should be considered (Maris et al., 2000). Most of the data is inconclusive and tainted by cultural and ethnic factors. Specifically, countries with more Catholic, Muslim, Protestant, Jewish, Buddhist, or Hindu followers report inconclusive suicide data despite the projection that their suicide rates should be low because of the high numbers of people committed to their faith. Yet, Hungary, whose Catholic population is 68%, has the second highest suicide rate in the world, and Austria, whose Catholic population is 90%, has the fifth highest suicide rate in the world. However, the Philippines, whose Catholic population is 80%, has a low suicide rate (65th), and Mexico, whose Catholic population is 90%, has a low suicide rate (54th). One would infer that religion alone does not prevent suicide; other contributing variables probably exist (Maris et al., 2000). In the United States, the most well established study on suicide and religion was conducted by R. W. Maris in 1981 at Johns Hopkins University (Maris et al., 2000). Known as the Chicago Sample Survey of Suicide, Maris (1981) compared the data on religion and suicide in Chicago to similar data from New York City. Overall results found that Protestants from Chicago and New York City have the highest suicide rates next to Catholics and Jews, 37 consecutively. Interestingly, Durkheim had predicted Jews would have the second highest suicide rate, not the third. In addition, the suicide rate of Protestant and Catholic males was twice as high as that for women, and in men over 65 years of age, the suicide rate exceeded females by a ratio of 3 to1 (Maris). Empirical evidence in the study by Maris (1981) suggests that a person’s commitment to adhere to the principles of a religion in some way protects that individual from seeking out suicide as a solution. Protestant males from Chicago and New York reporting little involvement with the church had the highest suicide rates; whereas, Catholic females from Chicago and New York reporting to be heavily involved with the church had the lowest suicide rates. Additional research by Early (1992) and Nisbet (1995, 1998) reported comparable data to that of Maris on African Americans suicidal ideation and authority of religion. Maris et al. (2000) surmised from Maris (1981) several points for consideration when interpreting data on suicide and religion. By and large, Protestants experience higher rates of suicide probably because their faith supports the idea that individuals are responsible for the perseverance to solve their own problems in order to reach salvation. Catholics believe they share this responsibility with God and to go against God and end life is sinful. Jews tend to share the responsibility of solving problems equally between individual and God, resulting in a possible reason why their suicide rates are lower than Protestants or Catholics (Maris et al., 2000). 38 On the whole, religious perspectives appear to have some impact on suicide rates (Maris et al., 2000). Globally, researchers deduced that religious integration, commitment, and networking have value and potentially could reduce suicide (Stack, 2001). Nevertheless, due to the myriad of cultural variables in societies, there seems to be an unclear causal connection between suicide and religion (Maris et al., 2000). Most religions today view suicide as a sinful act that defies the being they worship (Maris et al., 2000). Stack (2001) studied religious integration, religious commitment, and religious networks of support to find that these systems had the potential to improve a person’s life; thereby, the person avoids death by suicide. Evidence from studies conducted by Maris (1981), Early (1992), and Nisbet (1995, 1998) found that individuals engaged in religious practices were less likely to perceive suicide as an option. Summary Suicide is the act of an individual ending his or her life. It appears individuals seeking suicide believe ending their lives is the only resolution to their predicaments. As a result, much of the attention in the literature is focused on theories that attempt to understand why an individual might take his or her life. Prominent suicidologists with a background in psychology report that individuals contemplating suicide do so because they think suicide is the best way to relieve their psychological pain (Leenaars, 2003). Durkheim (1951) found that individuals disconnected from or disappointed with their social group appeared to be most at 39 risk for suicide. Finally, most religions view suicide as a sinful act (Maris et al., 2000). As such, studies demonstrated that individuals engaged in religious practices were less likely to consider suicide as a solution to their problems (Early, 1992; Maris, 1981; Nisbet, 1995, 1998). Suicide and College Students The proportion of college students turning to suicide as a solution to their problems has steadily increased over the last century (Haas, Hendin, & Mann, 2003; Magoon, 2000; Schwartz & Reifler, 1988; Silverman et al., 1997). As a result of this increase, more attention in the research is being paid to the rates at which college students attempt and commit suicide (Haas et al., 2003). This section provides: (a) an account of the suicide rates in college students during the last century, (b) a review of the Big 10 Student Suicide Study, and (c) an overview of suicidal ideation and suicide attempts in college students. Suicide Rates of College Students During the first half of the 20th century suicide rates for college students were relatively low and the cause of such incidents was usually associated with the stress of academic life (Raphael, Power, & Berridge, 1937). It was not until the late 1930s that researchers started to record college student suicide rate. Raphael et al. (1937) reported that five students from Ivy League institutions committed suicide between the years 1932 and 1936. This figure accounted for one half of the total deaths on campus (Raphael et al.). An overwhelming majority of these suicides were committed by males (Schwartz & Reifler, 1988). 40 Suicide gained more attention from researchers when the number of suicides tripled for men between the ages of 15 to 24 and doubled for women in that same age category by the end of the 1970s (Haas et al., 2003). Researchers still thought suicide was the result of intense academic pressures of college life (Haas et al.). Suicide rates at Ivy League institutions continued to gain attention during the 1970s. Cornell reported six suicides, the University of California at Berkeley had 23 suicides, and the University of Massachusetts at Amherst had twenty suicides (Schwartz & Reifler, 1988). When reviewing data from a multi- institutional sample of 117 institutions, 210 suicides occurred from 1971-1979 and this sample suicide rate is 7.3 per 100,000 students (Schwartz & Reifler). These data concur with another survey conducted by the American College Health Association Mental Health Annual Program Survey which reported a national college student suicide rate of 7.53 per 100,000 (Schwartz, 1995). Among college students, as in the population at large, men appeared to commit suicide more often than women (Schwartz & Reifler). Suicide became the second leading cause of death among college students in the 1980s (Haas et al., 2003). The rate of suicide remained steady during this decade. Research published by Lipschitz (1990) examined closely the limitations of the current studies on suicide rates of college students. Lipschitz believed the incidence of suicide on college campuses across the country was difficult to calculate. For example, suicide rates for college students ranged from five to 50 per 100,000. In addition, university administrators’ ability to define the 41 death of an individual college student was difficult. In many cases, administrators were unsure whether the student’s death was a suicide, an accident, or a homicide. Inconsistencies in the definition of suicidal actions, and the challenges administrators face in determining whether or not the student intended to commit suicide, have the potential to skew suicide data on college students. The Big Ten Student Suicide Study The most comprehensive research on student suicide was the Big Ten Student Suicide Study (Silverman et al., 1997). The purpose of the Silverman et al. study was to determine the college suicide rate by analyzing 261 suicides that occurred at twelve Big Ten campuses from 1980 to 1990. Variables considered in this study included age, gender, and race. Overall, results indicated that the rate of college student suicide was 7.5 per 100,000 students. This figure is consistent with findings by Schwartz (1995) and Schwartz and Reifler (1988). In addition, Silverman et al. (1997) identified the age group of individuals with the highest suicide rates. He found that males and females aged 20 – 24 represented 46% of the suicides. Females aged 17 – 19 represented 31% and males in that same age range represented 25% of the suicides. These data coincide with the research from the Center for Disease Control and Prevention (2006) which found that youth aged 20 – 24 commit suicide 68% more times than adolescents aged 15 – 19. The gap between the numbers of men versus the number of women completing suicide is not as significant as it had been in previous decades. Silverman et al. concluded that the completed suicide rate for 42 college students remained below the national average of 12.4 per 100,000. College students were not ending their lives as frequently as young people in the general population in part because students did not deal with life stressors associated with keeping a job and paying the household bills, plus students had easy access to mental health services on campus when in need of help (Silverman et al.). The rate of suicide on college campuses has tripled in the last 60 years and continues to be the second leading cause of death among college students (Center for Disease Control, 1995). The University of Maryland College and University Counseling Center Directors’ data bank reported 163 suicides representing 78 large universities and 85 small colleges (Magoon, 2000). Collectively these institutions represent 1,730,000 students, making the suicide rate 9.4 per 100,000 (Magoon). This rate of college student suicide is somewhat higher than the rate found in the Big Ten Student Suicide Study by Silverman et al. (1997). In another study conducted by Schwartz (1995), no differences in the rate of suicide were found at selective universities. Researchers project that each academic year approximately 1,088 students nationwide will end their life while attending college (Silverman et al., 1997; U. S. Department of Education National Center for Education Statistics, 2001). The odds of a student committing suicide are approximately 1000 to one (Appelbaum, 2006). Despite the fact that suicide is the second leading cause of death among college students, researchers determined that the college student 43 suicide rate is 7.5 per 100,000 individuals, below the national average of 12.4 per 100,000 persons (Silverman et al., 1997; U. S. Department of Education National Center for Education Statistics, 2001). Suicidal Ideation versus Suicidal Attempts The most recent data indicate that more students seriously think about committing suicide than those that actually attempt it (Barrios et al., 2000; Brener et al. 1999; Langinrichsen-Rohling et al., 2004; Westefeld et al., 2005). Suicidal behaviors studied in the literature are typically categorized in one of two categories: (a) college student suicide ideation, those students that seriously think about committing suicide; and (b) college students who actually attempt suicide. In this section, data on students reporting suicide ideation versus data on the rates college students attempt suicide will be reported and interpreted. Suicide ideation in college students. Multiple large scale studies on college student suicide ideation have been published. The American College Health Association (2001) surveyed 16,000 undergraduate students from 28 colleges and universities across the country and found that 9.5% of the students said they seriously contemplated committing suicide. These data are supported by two other national research studies by Brener et al. (1999) and Barrios et al. (2000). Brener et al. surveyed a national sample of 4,609 undergraduate students from two-year and four-year public and private colleges and universities and found that 10% of the respondents seriously thought about committing suicide. Barrios et al. used data from the National College Health Risk Behavior 44 Survey and calculated that of the 2,857 respondents, 11.4% seriously thought about committing suicide. A small study by Furr, Westefeld, Gaye, McConnell, and Marshall (2001) achieved similar results. About nine percent of the students participating in this study reported experiencing serious suicidal thoughts (Furr et al., 2001). In the last five years, the number of students seriously thinking about committing suicide has continued to rise. Two recent studies published by Langhinrichsen-Rohling et al. (2004) and Westefeld et al. (2005) found that more students are considering suicide as a serious option. In a survey of 383 students from a large southeastern university, Langhinrichsen-Rohling et al. reported that 22% of the student participants seriously considered committing suicide. Westefeld et al. found similar data to what was reported by Langhinrichsen- Rohling et al. Four large universities from the mid-west and south yielded 1,865 respondents, and 24% surveyed said they seriously thought about committing suicide (Westefeld et al.). Based on these percentages it appears that the number of students thinking about ending their life is on the rise. Yet, a study by Kisch et al. (2005) yielded data inconsistent with the Langhinrichsen-Rohling et al. and Westefeld et al. studies. Kisch et al. conducted a large scale study on college student suicide ideation involving 15,977 students from 28 campuses, and he found that only 9.5% of the respondents seriously considered committing suicide. This data supports earlier studies by Barrios et al. (2000), Brener et al. (1999) and Furr et al. (2001). 45 The studies mentioned above were composed of research samples of mostly traditionally aged college students who were asked to provide responses to a survey instrument in their introductory psychology class or social science course. The survey instruments used in these studies were not identical but the thematic content of the items related to suicidalogical topics is consistent among the survey instruments. Thematic content in these instruments included measuring opinions regarding suicidal behaviors, measuring proneness to suicide, measuring exposure to suicide, and measuring perceptions of suicide on a college campus (Langhinrichsen-Rohling et al., 2004; Westefeld et al., 2005). The data suggest that suicide ideation in college students appears somewhat inconsistent but could be viewed as increasing. Collectively, these studies reveal that the percentage of college students reporting suicide ideation has increased from about 9.0% to 22%. However, findings by Kisch et al. (2005) support those of early research by Barrios et al. (2000), Brener et al. (1999) and Furr et al. (2001) with 9.0% if the colleges students reporting suicide ideation. The apparent difference between 9.0% and 22% would lead one to assert that the data are inconsistent. College students attempting suicide. The American College Health Association, or ACHA (2004), Brener et al. (1999), Barrios et al. (2000), Langhinrichsen-Rohling et al. (2004), and Westefeld et al. (2005) also reported data on the number of students attempting suicide. Significantly lower than the suicide ideation data previously mentioned, these data are steadily increasing but 46 appear somewhat inconsistent. The ACHA (2004) found that 1.5% of their respondents attempted suicide, and two percent reported attempting suicide in the Brener et al. study. Yet, in the Barrios et al. study only 0.4% of the respondents said they attempted suicide. This figure increased significantly with the Langhinrichsen-Rohling study where they found that 9.1% of their participants attempted suicide, yet only 5% of the participants from the Westefeld et al. study reported attempting suicide and only 1.5% of the participants in the Kisch et al. (2005) study reported attempting suicide. These data confirm that the number of students thinking about suicide far exceeds those that actually attempt to take their life. However, the data measuring the number of students that attempt suicide is increasing but somewhat inconsistent; whereas the data measuring student suicide ideation is clearly on the rise. Suicide ideation and suicide attempts were also compared by gender (Barrios et al., 2000; Brener et al., 1999; Langinrichsen-Rohling et al., 2004; Westefeld et al., 2005). Specifically, Barrios et al. found that men and women were equally likely to report suicidal thoughts. Conversely, they revealed that men were more likely than women to complete suicide, and women were more likely to engage in a nonfatal attempt. Holmes and Holmes (2005) found similar results. They reported that men were three times more successful in completing suicide and women were three times more likely to attempt suicide (Holmes & Holmes). They reasoned that since men are seven times more likely to use a 47 firearm to commit suicide, and a firearm has a high rate of success, more men successfully commit suicide (Holmes & Holmes). Furthermore, Langhinrichsen-Rohling et al. (2004) mentioned that college men scored higher on the suicide proneness scale, a survey instrument used to identify behaviors linked to suicidal ideation, even though no significant gender differences in rating suicidal ideation were found. Westefeld et al. (2005), on the other hand, concluded that there were no significant differences in the way men and women thought about suicide, threatened suicide, or attempted suicide. For both genders, those who thought about suicide had a significantly higher risk of attempting suicide than the men and women who did not contemplate suicide (Westefeld et al.). Recent data on college student suicide ideation and college student suicide attempts demonstrate that behaviors related to suicide exist despite the inconsistencies in the data. Far more college students think about suicide than attempt suicide (Barrios et al., 2000; Brener et al., 1999; Langinrichsen-Rohling et al., 2004; Westefeld et al., 2005). College men and women show evidence of suicide ideation at about the same rates, but college men are generally more likely to attempt a fatal form of suicide (Barrios et al.; Holmes & Holmes, 2005). Summary Suicide is the second leading cause of death among college students; the college student suicide rate is 7.5 per 100,000 individuals, which is below the national average of 12.4 per 100,000 persons (Silverman et al., 1997; U.S. 48 Department of Education, 2001). Data on the numbers of college students experiencing some form of suicide ideation appear inconsistent. A summary of the research found that the percentage of students reporting suicide ideation on their campus varies from 9.0% to 22% (Brener et al., 1999; Furr et al., 2001; Kisch et al., 2005). Interestingly, college men and college women show evidence of suicide ideation at about the same rates, but college men are generally more likely to attempt a fatal form of suicide (Barrios et al., 2000; Holmes & Holmes, 2005). Deans of Students Within the Division of Student Affairs, DOS are primarily responsible for the development of the whole student in support of the academic mission of the institution (Blue, 1972; Eddy et al., 1988; Nuss, 2003; Sandeen & Barr, 2006). As manager, mediator, leader, and educator, DOS direct the programs and services that facilitate student development (Sandeen, 1991). As lead student advocate, DOS respond to students’ concerns (Komives et al., 2003). DOS possess the ability to influence student behavior regarding suicide through the structure of social integration systems that enhance student development (Stogdill, 1969). A review of the position, DOS, on college campuses follows. Development of Deans of Students The latter half if the 19th century saw the matriculation of an unprecedented number of college students for its time. Undergraduate matriculation tripled from 9,371 in 1869 – 1870 to 27,410 in 1900 (U. S. 49 Department of Education National Center for Education Statistics, 2008). Increasing numbers of students on campus required college administrators to monitor student behavior (Gerda, 2006). During this period, positions known as Dean of Women and Dean of Men emerged (Appleton et al., 1978; Gerda). Dean of women. The time period from post Civil War to well into the 20th century saw a significant increase in the number of women attending post secondary education institutions (Appleton et al., 1978; Schwartz, 2002). Enrollment of women accounted for nearly 47% of the total undergraduate population during the latter half of the 20th century (Schwartz, 2002). Single women living in a coeducational environment with single, young men generated a new set of social challenges for university administrators. It was during this time period that the Dean of Women position was established (Schwartz, 2002). Alice Freeman Palmer was credited with serving as the first Dean of Women in 1892 at the University of Chicago (Schwartz, 1997). A graduate of the University of Michigan and former president of Wellesley College, Palmer accepted this part-time dean’s position due in large part to her passion for women’s studies. She could not leave her husband in Boston so she settled for a part-time assignment as Dean of Women. President Harper agreed to hire Palmer’s friend and colleague, Marion Talbot, full-time to serve as Dean of Women for the University College in Palmer’s absence. Three years after Palmer’s appointment, Palmer resigned and Talbot became the full-time Dean of 50 Women and Assistant Professor of Domestic Sciences at the University of Chicago (Schwartz, 1997). New professionals on campus with no history or experience to draw from, the Dean of Women struggled to define their role on campus (Schwartz, 1997). Maintaining a high level of professionalism was thought to be the best course of action during this time period while they were trying to define their duties. The Dean of Women embraced the values of the academic world and believed their position and responsibilities as faculty made them a successful dean (Schwartz, 1997). As a result, the Deans of Women believed their growth and success as dean was dependent on the organization of professional affiliations. The first recorded gathering of the Dean of Women was organized by Dean Marion Talbot in 1903 (Gerda, 2006; Schwartz, 1997). Deans of Women from seventeen universities around the country attended this forum at the University of Chicago to discuss a variety of women’s issues including: housing, social etiquette, self- government, leadership, and athletics (Schwartz, 1997). From the dean’s perspective, these gatherings were the first step toward identifying student issues in order to enhance the advancement of this profession (Schwartz, 2002). As a result of these encounters, Dean Lois K. Mathews from the University of Wisconsin published The Dean of Women, the first book on the subject of being Deans of Women (Schwartz, 1997). 51 Initially members of the American Association of University Women (AAUW), the Dean of Women used this forum to develop their organizational skills, build their research, and start a graduate program on women’s studies (Schwartz, 2002). In 1916, they established a national group called the National Association of Deans of Women (Schwartz, 2002). This group focused specifically on assembling a professional organization focused on developing the skills and training required to be Dean of Women (Schwartz, 1997). Additionally, the women emphasized research and publication on the best methods of counseling students (Schwartz, 1997). Deans of Women emphasized the importance of research in order to advance their professional skills (Schwartz, 2002). They encouraged female scholars to attend graduate school, and they documented their activities and conducted research on their work as Dean (Schwartz, 2002). Most of the scientific research conducted by these women occurred at Teachers College at Columbia University (Schwartz, 1997). A graduate student from Teachers College, Jane Jones, published her study entitled A Personnel Study of Women Deans in Colleges and Universities in 1928. She found that 40% of the 263 deans that responded held the rank of full professor. Many of these faculty members taught in traditionally female departments like nursing and education (Schwartz, 1997). Representing a background in academics, it was logical to infer that a value in research and scholarship existed among these professional women. 52 Professor of Education at Teachers College, Ruth Strang, and colleagues Sarah Sturtevant and Esther Lloyd-Jones led the Committee of the National Association of Deans of Women research efforts (Schwartz, 1997). They are credited with identifying the philosophical and scientific aspects of deanship. The philosophical aspects represented the experiences that inspired women to the dean’s position and the scientific aspects of the research focused on experimental studies that identified best practices when working with students (Schwartz, 1997). The results of this research were published in an academic journal founded by the dean of women, and they went on to publish books on how to be a dean as early as 1916 (Schwartz, 2002). Schwartz (1997) observed that Deans of Women were committed to legitimizing the Dean’s position within university circles as demonstrated by their efforts to conduct research, publish books, and establish an academic journal associated with the profession, Dean of Women. Yet, Deans of Women reported the task of legitimizing this profession difficult in a community dominated by men. Most of the significant accomplishments made by these women were ignored by the academy (Appleton et al., 1978; Schwartz, 1997). Additionally, Schwartz (2002) identified the most prominent difference between the Dean of Men and the Dean of Women was their stance on academics as it related to the profession. Dean of Men valued and nurtured the interpersonal skills of the men seeking this position, and the Dean of Women sought to enhance their professional development through scholarship (Schwartz, 2002). 53 Dean of men. Not long after Dean Palmer’s appointment to Dean of Women in 1892, university presidents began to search for a male member of the faculty who demonstrated strong interpersonal skills and the ability to reach students to serve as dean to the male student body (Schwartz, 2002). In reality, according to the early deans of men, their president needed an administrator to deal with student behavior matters. University presidents called on a member from the faculty who appeared to share a kinship with students to fill the role of Dean of Men. College administrators agreed that simply having the communication skills needed to interact with students was not much of a professional job description, but it appeared to be a good place to start (Schwartz, 2002). During his tenure as Dean of Men in the 1900s, Thomas Clark from University of Illinois recalled that he had no job description, and loosely defined himself to be college disciplinary officer (Clark, 1921). His president seemed only concerned that someone took care of student problems (Clark). President Charles Eliot from Harvard University recalled feeling the burden of an increased work load that left no room for handling student problems, and as a result, he hired two faculty deans; one for the faculty and one for the students (Appleton et al., 1978; Schwartz, 2002). Credited with being one of the first administrators to hire a Dean of Men, President Eliot selected the very popular Professor Le Baron Russell Briggs to serve as Dean of Men. Like Briggs and Clark, most of the early deans were chosen to serve as Dean of Men not for their professional qualifications, but instead, for their endearing personalities and 54 ability to connect with students. Dean Scott Goodnight from the University of Wisconsin, Dean Robert Rienow from Iowa University, Dean F. F. Bradshaw from the University of North Carolina, and Dean Stanley Coulter from Purdue University were among the first deans to assert their role as in loco parentis to the young men attending their institution (Schwartz, 2002). Faculty performing the duties of Dean of Men had several professional philosophies in common. First, with overwhelming passion, it was thought that a dean of men was born not made (Schwartz, 2002). Skill in creatively guiding men to be responsible citizens could not be taught. It was an inherent ability, a gift, a calling. No graduate education, research, or training was necessary to be a Dean of Men; encouraging faculty to apprentice with an established dean was considered the best form of training (Schwartz, 2002). These beliefs are in contrast to those of the Dean of Women. Second, across the board, deans agreed that their function on campus was undefined, unclear, and offered little expectation (Schwartz, 2002). No job descriptions existed, no professional standards existed to emulate, and many university presidents and trustees had no idea what a Dean of Men was supposed to accomplish. When Dean Stanley Coulter from Purdue University approached his board and asked the board to outline his duties as dean of men; reluctantly, they had no answer for him. Dean Coulter searched out his colleagues for advice and direction only to find that most deans were the 55 president’s dumping ground for the problems for which the president had little patience (Schwartz, 2002). Third, early deans operated as university disciplinarian officers. Coping with academics and new found social freedom proved to be a challenge for young men in college (Schwartz, 2002). These young scholars were well intended souls but lacked discipline and required guidance from faculty as recalled by the early deans. It was their vision to teach young men to be honorable, generous citizens. Achieving this goal would happen by revealing the realities of the world to students in a somewhat controlled environment, permitting students to learn under the guidance of the dean (Schwartz, 2002). Teaching citizenship and good behavior was a huge undertaking for the deans. To facilitate their professional growth, the deans met informally as a group, for the first time in 1919, to swap stories and solutions with one another. Meeting minutes from the deans’ first informal gathering mentioned discussions of students’ behavior, fraternity initiations and finances, students’ affiliation with landlords, classroom attendance, and student government (Turner, 1936). A year later, the deans met again to formalize the function of their group. They founded the National Association of Deans of Men, whose primary role was to define the character of deans of men. Senior deans like Clark shared what became known as oral histories at annual meetings with the hope that his leadership would benefit the less experienced deans (Turner). Soon after, Clark (1921) published 56 his most remarkable disciplinary cases in his book, Discipline and the Derelict: Being a Series of Essays on Some of Those Who Tread the Green Carpet. Deans of Men were academic scholars revered for their abilities to make meaningful connections with students with the goal of teaching and reinforcing responsible citizenship (Schwartz, 2002). University presidents offered no job description or professional standards to the Dean of Men; instead, the dean took initiative to define the purpose of his position while serving as university disciplinarian officer. Finally, Deans of Men were viewed to have been born for this leadership position not trained to fill it (Schwartz, 2002). Overall plight of deans. In an analysis of the accomplishments of the Deans of Women and Deans of Men, Appleton et al. (1978) were unsure why these professionals were judged poorly by their academic peers. A review of their accomplishments revealed that deans were generally compassionate, warm, and optimistic professionals. They exhibited characteristics that indicated they felt respect and concern towards the students. They outlined their duties with little guidance from the college president, gained support for the value of student services with little financial support, and met with their colleagues to form their purpose and values. These leadership qualities along with a liberal arts education were thought to be a formula for a successful dean. Unfortunately, Appleton et al. concluded that attempts made by Deans of Women to present institutions with the foundation of the future work in student affairs lacked the credit it deserved. 57 One Dean of Students After World War II, university administrators dismantled the traditional model of Dean of Men and Dean of Women and replaced these positions with one DOS (Schwartz, 2002). The Dean of Men survived this re-organization despite the fact they did not embrace the research philosophies coveted by the women and academe. Nevertheless, veteran Dean F. F. Bradshaw from The University of North Carolina admitted that student personnel workers, despite their kinship with students, benefited from research in the best practices for academic advising and counseling (Schwartz, 2002). The role of DOS was similar to that of the Deans of Men and Women. The development of the whole student through the establishment of student life services that improved students’ social/emotional development and abilities to be a productive citizen remained intact as the main responsibility of DOS (Eddy & Sharma, 1986). As outlined by Barr and Keating (1985) the role of DOS included: serving as “in loco parentis” to students in the absence of their parents; the development and implementation of a campus crisis management plan; and program development within student services that addressed students interests and university traditions. However, even though these responsibilities of the DOS remained constant, it is important to note that services in student life had expanded significantly. Student life services led by DOS usually included: Admissions and records offices; orientation; counseling and health services; student union 58 activities; student financial aid; career centers; housing offices; academic advising centers; and diversity offices. For the first time, these services were located in separate units across campus in order to meet the individual needs of the students (Eddy & Sharma, 1986). Collectively, these campus service offices typically reported to DOS (Eddy & Sharma). Post World War II lead student affairs personnel were known as DOS (Schwartz, 2002). The use of the title Dean of Men and Dean of Women was rendered obsolete, and the role of DOS was predominantly filled by men. Position responsibilities of the DOS remained somewhat the same and showed signs of expanding. In addition to serving in the role of “in loco parentis” and responsible for the development of the whole student, post World War II DOS were also in charge of student services like student health, financial aid, and housing services, to name a few. Research and documentation of student activities and interactions continued to be a priority (Schwartz, 2002). A Dean’s Perspective Post World War II Little empirical evidence is available about the administrative role of DOS. In its place, DOS themselves have recounted in published documents their understanding of the role of the DOS. The Peabody Journal of Education cites an article written by Ronald E. Barnes, the DOS at Iowa State University in the 1960s. An account of the job duties performed by Barnes (1963) as DOS is presented. 59 Barnes (1963) recalled that normal hours and routine on the job as DOS did not exist. Instead, he described how he spent his days responding to the latest crisis or event like an unscheduled student gathering, calls from worried parents, meetings with disgruntled landlords, and students that needed to speak with an administrator about the latest pressing matter. In addition to handling the current crisis, the DOS carried a full social calendar. Students usually requested his or her presence at evening dinners, speaking engagements, and dances. This reactionary work style did not leave much time for professional development (Barnes). According to Barnes (1963) the role of DOS was the university “middleman” (p. 4). Granted the unofficial title of college disciplinarian, DOS attempted to balance the enforcement of the rules with the students and the student life related demands of the faculty. In this position, DOS were labeled the “controversial figure” on campus (p. 4). Students viewed the DOS as someone who was trying to control them, and faculty seemed to have little respect for the DOS because he or she was not a member of the academic family. Consequently, students were not sure if they could trust the DOS, and faculty thought the DOS was an easy target to cast blame on when an event at the college failed. All things considered, administrators accepted the DOS in order to avoid getting involved with the DOS undesirable work load. The campus community had no interest in the problems encountered by the DOS. Essentially, DOS had minimal support on campus, little direction from authority, and an 60 undefined role. Barnes (1963) went so far to say, “…becoming a dean of students is, for all but a few, a dead-end street to higher university positions” (p. 4). Dealing with student problems was identified by Barnes (1963) as the most undocumented issue in the literature. Little guidance on how to manage student crisis was available and he was left to manage these problems alone. On a typical day, Barnes dealt with a myriad of issues related to student concerns. For example, he responded to court orders filed by students who alleged that their due process rights were violated. He met with members of student government when they demanded college rules be changed. He answered the members of the racial equality committee when they demanded action be taken against landlords and the director of housing for practices of discrimination. Plus he acted in response to the president’s demand that he fire the editor of the school newspaper, quickly and quietly. These incidents were a sample of the numerous challenges identified by Barnes that said crossed his desk daily. Finally, DOS deal with matters related to problems with personnel, a commitment to numerous committee assignments, and attention to an administrative work that includes endless paperwork documenting campus events and interactions with students, generating reports related to matters about policy and procedures, and writing students’ letters of recommendation. An attempt to find time to prepare a budget and read relevant research articles on student personnel were additional important goals (Barnes, 1963). 61 Despite the challenges, Barnes (1963) recounted the many advantages to this position, the number one advantage being that DOS was the opportunity to make a significant contribution to the life of a student. DOS are directly involved in the development of students which yields a personal and professional sense of accomplishment. The day-to-day interaction with students has the potential to positively influence the students’ emotional and cognitive maturity. Second, the DOS of students supports and facilitates students’ engagement with the campus community. The DOS authority has the potential to assist students with becoming active campus participants. Third, the DOS collaborates with faculty on programs that support student development. Ultimately, Barnes recalled that he had “…the opportunity to stimulate others to improve their services” (p. 7). In the end, Barnes (1963) viewed the DOS as a university “Middleman” and campus disciplinarian. He summed up the role of a DOS with the following statement: A dean of student’s primary responsibility is to develop attitudes among colleagues and students which encourages and enables each to increase his understanding and appreciation of the other’s position, needs, and goals. He should be the primary representative and spokesman through whom the faculty and students communicate with each other officially on matters involving the welfare of students outside the classroom. The degree to which effective communication develops is largely dependent upon the respect accorded the dean by those within the academic 62 community. He can play a most important role in minimizing misinformation and misunderstanding while promoting an atmosphere of mutual trust and respect (Barnes, p. 9). These lofty expectations of DOS were not necessarily the reality on the job as explained by Barnes (1963). He reported that students believed he was trying to control them in his capacity as campus disciplinarian, and faculty failed to respect him because he was not a member of the academic academy. Thus, the DOS was an easy target to cast blame on when matters of the wellbeing of a student were at stake. However, Barnes believed his opportunities for making a significant difference in a young person’s life while managing campus civic engagement activities far outweighed the stresses of the job. Authority of Deans of Students The research attempting to analyze DOS power appears to be consistent over the last three decades. Appleton et al. (1978), Komives et al. (2003), Nuss (2003), Sandeen (1991), and Sandeen and Barr (2006) believe that DOS formal authority is given, earned, and presumed. Authority is given to DOS by the person who hired him or her; in most cases the individual presenting the authority to the DOS is the university president or chancellor or chief academic officer such as the provost. Authority is earned when DOS job performance is deemed competent by the campus community; power is present when the DOS demonstrates sound judgment and maturity when making decisions. Authority is presumed because DOS possess the power to make decisions as they see fit; 63 this latitude of authority is a direct result of DOS involvement in solving the problems related to student life (Appleton et al.; Komives et al.; Nuss; Sandeen; Sandeen & Barr). The following paragraphs further explore DOS authority given, authority earned, and authority presumed. First, DOS are given authority, in part, to act in the best interest of the institution (Appleton et al., 1978; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). Administrative support is considered crucial to the DOS. The ability to respond to university crises requires more than the authority granted by the president or board. DOS authority is boosted by the cooperative efforts of multiple student affairs personnel that work for the DOS. The backing of external constituencies like alumni and community leaders increases DOS given authority. These collective bases of support enhance DOS authority in an effort to achieve success for the institution (Appleton et al.; Komives et al.; Nuss; Sandeen; Sandeen & Barr). Second, DOS authority is earned by the quality of DOS performance. In many instances, power earned by the worth of DOS accomplishments far outreaches DOS given authority by virtue of his or her position of power (Appleton et al., 1978; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). In particular, Appleton et al. noted that earned authority was largely dependent on DOS talents to deliver on their potential. In doing so, DOS earn authority because of “a reputation of excellence, of sound judgment, or maturity, of winning battles with minimal costs to others’ egos…” (Appleton et al., 64 pp. 59-60). DOS earn authority by exercising good reasoning in order to craft competent decisions (Appleton et al.; Komives et al.; Nuss; Sandeen; Sandeen & Barr). Third, DOS authority is presumed in large part by the institutional hierarchy, and DOS ill-defined scope of responsibility (Appleton et al., 1978; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). DOS are required to assume authority and take responsibility for solving problems related to student life. In turn, responding to university problems means DOS conduct is led by what they think is right. Defining what is right is an opportunity of demonstrate competent decision making which ensues a strong sense of presumed authority for DOS (Appleton et al.; Komives et al.; Nuss; Sandeen; Sandeen & Barr). The assumptions of DOS authority infer that authority is given, earned, and presumed (Appleton et al., 1978; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). Appleton et al. stated, “….emphasize the importance of power derived by competent performance; …the authority to act is not limited to the formal offer of authority from a superior. Colleagues, students, other administrators, and staff for whom the DOS is responsible, as well as external constituencies, all contribute to the support necessary for successful deaning” (Appleton et al., p. 64). Institutions grant the DOS substantial power based on position of authority and scope of responsibilities. However, DOS ability to successfully meet the needs of the university body might 65 impact the DOS source of power (Appleton et al.; Komives et al.; Nuss; Sandeen; Sandeen & Barr). Deans of Students Today Contemporary literature on DOS has moved away from evaluating the formal authority of DOS due to the relative consistencies in its findings to analyzing the titles and the labels of this position. As well, studies examining the gender of these professionals are of special interest. An investigation of this contemporary literature is worthy of discussion because despite the changes in employment titles, labels, and gender issues the role and responsibilities of DOS has not changed drastically. In addition, it is unknown how these changes in employment titles, labels, and gender issues will impact DOS’ perceived position of authority. Research on the titles and the labels of lead student affairs’ administrators determined that leaders in student affairs held a variety of titles such as: vice president, DOS, or director; and their label included: student affairs, student services, student development, or campus life. In addition, the advancement of women in positions of authority within student affairs “has more than doubled overall from 22% in 1984 to 45% in 2006 (Tull & Freeman, 2008, p. 272). An understanding of the assortment of titles and labels along with insight into the substantial increase of women serving as lead administrators in student affairs merits research in these areas to determine how these changes impact DOS formal authority. 66 Tull and Freeman (2008) replicated Rickard’s (1985) study when they examined the collection of lead student affairs officers’ titles and labels in association with gender. Data for this study, and the previous study by Rickard, was obtained from the Higher Education Publications Directory. Data included: gender, institutional type/size, employment title, and division/department label. The analyses of data published from Rickard were used to compare with the findings noted by Tull and Freeman. Standardization of titles and labels for lead administrators in student affairs appears to be almost non-existent. Student affairs administrators report mixed opinions about the uniformity of working titles and labels (Tull & Freeman, 2008). Some believe standardization would provide more consistency thus lending more credibility to the field as a profession. Others believe the realities of institutional autonomy and diversity are an obstacle in reaching such consistency. Next, there appears to be an increase in the use of the title, vice president, among institutions with populations larger than 5,000. The authors contend that the size and diversity of larger institutions fit the organizational and governance structures typical of the title, vice president. In addition, many of these same larger institutions reported lead student affairs administrators with dual titles like vice president/DOS. This trend to assign dual titles to student affairs administrators is a strategy employed to streamline the organizational structure of personnel titles and department labels (Tull & Freeman). Finally, the most interesting finding according to Tull and Freeman (2008) 67 was the closing of the gender gap amongst professionals employed as lead student affairs administrators at all institutions who participated in this study. Women are entering the profession of student affairs at a quicker rate than men; Tull and Freeman anticipate that women employed as administrators in student affairs will soon outnumber the men who occupy positions as lead student affairs administrators. Similar findings were found in research conducted by Pickering and Calliotte (2000) where they found the percentage of women occupying a position in student affairs was 63%. Summary The establishment of Student Affairs is rooted in the belief that student affairs personnel are responsible for the development of the whole student in support of the academic mission of the institution (Nuss, 2003). The drive behind the vocation of student affairs grew out of support for the diversity of the students it serves. Ultimately, student affairs professionals are held accountable for designing a college community that maximizes the students’ experiences (Sandeen & Barr, 2006). Typically, DOS are the student affairs administrators in charge of the preparations to develop the whole student on university campuses in accordance with the university’s mission. Early on, DOS were known as Dean of Men or Dean of Women. Although they sought to serve as the care taker of students, each professional achieved that goal in a different manner. Dean of Men believed being dean required “on the job” training with little emphasis on research and 68 planning; Dean of Women believed the foundation of their success was their scholarly work related to being a dean (Schwartz, 2002). At present, post World War II Deans of Students subscribe to the views of the Dean of Men and Dean of Women. Experiential training and scholarly activity became the professional qualifications for current DOS. In terms of authority, DOS were often described as the university “middleman” and campus disciplinarian (Barnes, 1963). Appleton et al. (1978), Komives et al. (2003), Nuss (2003), Sandeen (1991), and Sandeen and Barr (2006) agree that DOS authority is given, earned, and presumed. Presidents grant DOS substantial power based on their position as student affairs administrator and range of responsibility in the development of the whole student. DOS ability to successfully meet the needs of the university body has the potential to impact DOS source of power (Appleton et al.; Komives et al.; Nuss; Sandeen; Sandeen & Barr). Finally, the gender gap between the numbers of men and the numbers of women serving as DOS is closing. It is unknown how this closing gap might impact perceived power and authority of DOS. Conceptual Framework DOS are known to be the student affairs administrators serving as direct university advocate and care-taker of undergraduate students (Komives et al., 2003). DOS play an important role in being responsible for the safety and well- being of students which includes finding the best methods for identifying students who might commit suicide (The Jed Foundation, 2006). The literature suggests 69 DOS possess the authority to handle such responsibilities. However, it is unknown whether or not there is an association between DOS’ opinions about suicide and their position of authority. Thus, the conceptual framework used for this study is twofold. First, the three factor model of attitudes toward suicide – acceptance, condemnation, and prevention is used to identify DOS attitudes toward suicide (Salander-Renberg et al., 2008). Second, the six base social power typology – coercive, expert, informational, legitimate, referent, and reward, classified by French and Raven (1959, 1968) is used to identify the power sources utilized by DOS. An overview of the research for this conceptual framework includes: (a) an investigation of individuals’ attitudes towards suicide, (b) the methods used to measure attitudes toward suicide, (c) a discussion of social power, and (d) the methods used to measure the social power bases. Individuals’ Attitudes toward Suicide Individuals’ attitudes and perspectives are central to the prediction of their intended behaviors; as such, research focusing on understanding individuals’ attitudes and perspectives about suicide attempt to measure opinions about suicide (Ajzen, 1985). The research acknowledges that studying attitudes toward suicide is complex and unpredictable (Bayet, 1922; Gibbs, 1968; Platt, 1989; Sale et al., 1975). In addition, individuals’ perceptions of suicide might impact their responses to surveys that attempt to measure attitudes toward suicide (Ginn, Range, & Hailey, 1988). Thus, present research continues to evaluate 70 attitudes toward suicide and discusses survey instruments which attempt to quantify individuals’ attitudes toward suicide. In this section, an overview of the research on attitudes toward suicide follows. A review of the literature on the study of attitudes toward suicide appears to center on exploring the associations between suicidal behaviors, suicide rates, and attitudes toward suicide of individuals in the general population (Salander- Renberg et al., 2008). The initial investigations studied the relationship between individuals’ attitudes toward suicide and suicide behaviors were conducted by Bayet (1922). His analysis yielded two attitudes toward suicide. First, individuals who condemn suicide support the notion of “moral simple” and believe the act of suicide to be morally wrong under any circumstance. Second, individuals who possess no absolute judgment for or against suicide support the notion of “moral nuance” (Bayet). Moreover, people subscribing to “morale nuance” believe that individuals should examine the motivations or circumstances for the suicide because they assert suicide might be acceptable (Pickering, 2002). Needless to say, Bayet found that the “morale nuance” individuals had higher rates of positive relationships of acceptance of suicide than the “morale simple” individuals. Gibbs (1968) investigated the relationship between individuals’ attitudes toward suicide and rate of suicide in their town. He discovered that people living in locations with high rates of suicide had predominantly intolerant attitudes toward suicide, and people living in locations with low rates of suicide were more accepting of suicide. Of course, individuals known to attempt suicide reported a 71 permissive attitude toward suicide. Outcomes from Sale et al. (1975) and Platt (1989) support the findings of Gibbs. Both studies concluded that people living in locations with high rates of suicide had predominantly intolerant attitudes toward suicide, and people living in locations with low rates of suicide were more accepting of suicide. The early efforts to study attitudes toward suicide were not without unanswered questions and limitations. According to Gibbs (1968), the direction of the individual’s influence regarding attitude and behavior was unclear. Namely, did the individual’s attitude toward suicide influence the individual’s behavior to commit suicide, or did the individual’s behavior to commit suicide influence the individual’s attitude toward suicide. In either case, it appears early researchers identified associations between attitudes toward suicide and suicidal behaviors and suicide rates. Present research on the study of attitudes toward suicide finds some individuals possess an intolerant view of the act of ending one’s life. McLaughlin (2007) studied health care providers’ attitudes of suicidal patients. A patient’s act of suicide elicited negative attitudes and behaviors from their health care providers. For example, the health care providers stated that para-suicidal patients did not really try to kill themselves, and they went on to rationalize the suicidal behavior of their patients as manipulative and attention seeking. McLaughlin asserts that these negative attitudes towards suicidal patients could negatively impact the patients’ treatment. 72 Morgan, Evans, Johnson, and Stanton (1996) also assessed health care providers’ attitudes toward suicide. In this study, health care providers participated in two lectures on suicide prevention. Researchers compared their attitudes toward suicide prevention pre-lecture and post-lecture. Results indicated that one-third of the health care providers reported possessing negative attitudes toward suicide prior to the lecture on suicide prevention. Interestingly, the post lecture on suicide prevention yielded a significant reduction in negative attitudes toward suicide prevention in the health care providers. The authors deduced that a better understanding of suicide prevention might reduce health care providers’ negative attitudes toward suicide (Morgan et al., 1996). People’s negative attitudes toward suicide have the potential to extend beyond the individual level to the community level. Ginn et al. (1988) believe the community attitude regarding suicidal behavior is to ignore it. According to the authors individuals displaying suicidal behavior were thought to be handled privately as a family matter. Based on these findings, Ginn et al. believed community awareness of suicidal behaviors should be discussed openly. Present studies of individuals’ attitudes toward suicide seem to reveal that individuals display negative thoughts and feelings about suicide (Ginn et al., 1988; Morgan et al., 1996; McLaughlin, 2007). Furthermore, persons negative attitudes about suicidal individuals extend to the community in the way communities tend to disregard the prevention of suicide (Ginn et al.). However, Morgan et al. found that educating individuals about suicide reduced the negative 73 attitudes individuals had toward suicide. Yet, Salander-Renberg and Jacobsson (2003) reported that populations with a permissive attitude toward suicide saw more suicides in their communities and more notably from young women. One might conclude that individual’s attitudes toward suicide seem to lack consistency in the literature. Measuring Attitudes toward Suicide Early research on suicide focused on attempting to understand why an individual might commit suicide. Theorists who subscribe to the psychological, sociological, or religious perspectives related to the act of suicide have attempted to define and analyze suicide from that perspective as discussed earlier in this review of the literature. However, an effort to measure individuals’ opinions and attitudes toward suicide is a new undertaking for researchers. In this section, a chronological overview of the instruments used to measure attitudes toward suicide starting with the Suicide Opinion Questionnaire (SOQ), the Suicide Attitude Questionnaire (SUIATT), and the questionnaire on Attitudes toward Suicide (ATTS), the instrument to be used in this study, is discussed. The suicide opinion questionnaire. The first researcher to design and implement a quantitative survey instrument for the purposes of measuring individuals’ attitudes toward suicide was Domino (2005). A recollection of the story behind the creation of the Suicide Opinion Questionnaire has become the foundation for current instruments that attempt to measure attitudes toward 74 suicide. Thus, recounting the development of the Suicide Opinion Questionnaire is warranted. In 1975 two undergraduates asked Dr. Domino to sponsor the implementation of a suicide hotline in South Tucson which was a predominantly Hispanic community with almost no mental health support services (Domino, 2005). The students found an old store front for their hotline service where Dr. Domino oversaw their efforts in addition to finding the students a couple of local psychologists to serve as consultants. At the end of the first term, the students had received few calls from the community. No suicidal individuals called the hotline. In light of these events, Professor Domino encouraged the students to consider doing research on the topic of attitudes toward suicide. Much to their surprise, few studies investigated individuals’ attitudes toward suicide (Domino). Knowing there was little information in the literature about attitudes toward suicide; Professor Domino found an opportunity to apply his knowledge of psychological testing to the study of attitudes toward suicide (Domino, 2005). He and several of his graduate assistants identified and documented attitudinal type statements from the articles they read on suicide. They compiled a total of 3,000 attitudinal type statements. Using content analysis they eliminated duplicate statements, and revised statements as needed. This effort yielded 138 questions about attitudes toward suicide. This initial survey instrument was administered to 96 undergraduate students twice during a six week time frame. According to Domino, “The 100 items with the highest test-retest reliabilities (all above r = 75 0.68) were kept and the Suicide Opinion Questionnaire (SOQ) was born” (p. 108). Attempts to determine the reliability and validity of the SOQ ensued. Domino, Gibson and Poling (1980) conducted a content analysis on the responses obtained from 800 college students from nine universities. This effort generated nine content areas, “…items that related to mental illness, religion, suicide as a cry for help, personal values, family aspects, motivational themes, demographic variables, incurable disease, and miscellaneous concerns” (Domino, 2005, p. 109). Domino, Moore, Westlake and Gibson (1982) used this data and went on to publish a study that yielded 15 factors which were found to be reliable and meaningful. In an attempt to improve the reliability and validity of the SOQ, others continued to generate measurable content factors through research utilizing the SOQ. Anderson, Lester, and Rogers (2008) identified several limitations of the SOQ. These authors dispute the effectiveness of the SOQ as a psychometric measure of attitudes toward suicide (Anderson et al., 2008). Domino (2005) refutes this position by stating that the SOQ was not developed to be used as a psychometric measure, but instead, was simply an instrument used to assess attitudes toward suicide. Another limitation of the SOQ is the old age of this instrument which might impact the manner in which respondents interpret the items on the instrument. Also, lack of clarity between items that are defined as attitudinal versus items that are defined as factual only added further disorder to 76 this instrument. The structural validity has been found to be questionable which might have an impact on the usefulness of the SOQ (Anderson et al.). Suicide attitude questionnaire. Diekstra and Kerkhof (1998), along with Domino et al. (1980), acknowledged the uncertainty of the theoretical and psychometric value of the SOQ. Furthermore, Diekstra and Kerkhof believed that theories about attitudes toward suicide and the measurement of attitudes toward suicide were overlooked in the item and scale construction of the SOQ. Thus, in order to address the aforementioned criticisms of the SOQ and respond to the need for empirical research on the relationship between attitudes toward suicide and suicide rates of various social groups, Diekstra and Kerkhof designed the Suicide Attitude Questionnaire (SUIATT). Diekstra and Kerkhof (1998) searched for a model they could use to classify respondents’ attitudes toward suicide. The authors determined that the consequences of committing suicide appeared to impact three distinct categories or referents of individuals: (a) actor, the individual committing suicide; (b) recipient, close friends or relatives of the individual that committed suicide; and (c) bystanders, people who knew the individual who committed suicide. Therefore, individuals’ attitudes toward suicide might differ depending on their relationship with the person thinking about committing suicide. Considering the attitudes of the various referents led Diekstra and Kerkhof to include survey items for actors, recipients, and bystanders. Definitions of attitudes were classified using Secord and Backman’s (1974) three component model of attitude: (a) 77 affective, emotional feelings about suicide; (b) cognitive, knowledge of suicide; and (c) instrumental, opinions about the act of committing suicide (Diekstra & Kerkhof). Findings of these initial efforts produced a survey instrument with 133 items using a five point Likert scale; items perceived to be negative attitudes were scored on the lower end of the five point scale (Diekstra & Kerkhof, 1998). The purpose of this questionnaire was to measure the respondent’s attitudes (affective, cognitive, or instrumental) toward suicide in relation to the respondent as an actor, recipient, or bystander. A group of 85 clinical psychology graduate students completed the 133 item instrument. Completion of a factor analysis eliminated 70 items and the remaining items were reformulated with each component of attitude having at least fifteen items in it. The improved 63 item questionnaire was then administered to 712 persons from eleven urban and six rural communities from the Netherlands (Diekstra & Kerkhof). Results of this study generated several interesting findings. Diekstra and Kerkhof (1998) found a significant difference (F = 35.19, p < .001) when running an analysis of variance on factors scored as “referents”, the differences between responses as an actor, recipient, or bystander. Interpretation of these results determined that the internal validity of this instrument is sufficient. Interestingly, when using a principle component analysis the authors found that the internal structure of the SUIATT uncovered that a person’s desire to commit suicide under physical or social stress was rather detached from the person’s attitudes 78 about how mentally troubled he or she should be in order to have the right to commit suicide. As a result, the authors believe future research should consider looking at the cognitive and instrumental attitudes toward suicide for differences between physical and social conditions (Diekstra & Kerkhof). Limitations of the SUIATT are documented by the authors and others in the literature. Diekstra and Kerkhof (1998) and others agree that the present construction of the SUIATT is complicated, long, and too time consuming (Etzersdorfer, Vijayakumar, Schony, Grausgruber, & Sonneck, 1998). An instrument that measures attitudes toward suicide in short form is needed. The long length of this questionnaire offers a wide range of subscales, 19 to be exact. Etzersdorfer et al. noted, “These subscales try to deal with very detailed aspects (e.g. emotional aspects, rationality, consequences for oneself and for others), which we found difficult to differentiate…” (p. 105). For example, Etzersdorfer et al. (1998) observed that items labeled emotional attitudes could just as easily been labeled as rational attitudes items making it difficult to clearly define the subscale categories. The opportunity to address the limitations of the SOQ and the SUIATT led Salander-Renberg and Jacobsson to design a questionnaire that measures attitudes toward suicide. Attitudes towards suicide questionnaire. The development of the Attitudes towards Suicide (ATTS) questionnaire sought influence from the authors who designed the SOQ and the SUIATT (Salander-Renberg & Jacobsson, 2003). While constructing the ATTS, Salander-Renberg and Jacobsson looked to 79 address the criticisms of the SOQ and the SUIATT, specifically the long length of these instruments. The purpose of the ATTS is to measure people’s attitudes toward suicide using a small scale survey instrument that could be completed by individuals in the general population. Of special interest to the authors was the study of the associations between attitudes toward suicide and suicidal behaviors of these citizens in the general population. Since this study is not focusing on the link between attitudes toward suicide and suicidal behaviors, this review will only highlight the findings on measuring attitudes toward suicide. Salander-Renberg and Jacobsson (2003) designed and tested two research studies using the ATTS over a 10-year time period from 1986-1996. Of primary concern to the authors was the need to design a short survey instrument that could be answered by the general population without negatively impacting the reliability and validity measures. The article, Development of a Questionnaire on Attitudes towards Suicide (ATTS) and Its Application in a Swedish Population, by Salander-Renberg and Jacobsson offers an overview of this decade long research endeavor whose focus was the study of individual’s attitudes toward suicide using a large population (see Appendix A). This review details the construction of the ATTS and the findings cited by Salander-Renberg and Jacobsson. The composition of the initial survey instrument, in the mid-1980s, yielded a 62 item questionnaire which utilized the three component attitude model: affective, cognitive and instrumental, the same one used by Diekstra and Kerkhof 80 (1998) in the SUIATT (Salander-Renberg & Jacobsson, 2003). To measure respondents’ attitudes toward suicide on the ATTS, the authors applied a 5-point Likert scale: agree completely, agree to a large extent, doubtful, do not agree, and do not agree at all, similar to the Likert scale used by Domino et al. (1982) in the SOQ. Randomly selected individuals, ages 18-65, from the general population were invited to complete the ATTS. A sample of 522 individuals completed the instrument, a response rate of 75 % (Salander-Renberg & Jacobsson). Findings obtained eight factors of attitudes toward suicide: permissiveness, unpredictability, incomprehensibility, non-communication, right to prevent, preventability, relation-caused, and suicidal process (Salander- Renberg & Jacobsson, 2003). This eight factor model accounted for 63% of the total variance with internal consistency rating varying from 0.66 to 0.34. Highest acceptable variance and internal consistency rates were found for attitude factor items categorized as permissiveness, unpredictability, and non-communication. As a result, Salander-Renberg and Jacobsson, “…decided to emphasize and focus on preventive aspects instead of cognitive aspects, leading to the exclusion of items primarily concerning causes of suicide” (p. 54). Ten years later, in 1996, Salander-Renberg and Jacobsson repeated their study of measuring attitudes towards suicide using the ATTS (Salander-Renberg & Jacobsson, 2003). The second attempt at this research emphasized finding ways to increase the reliability and validity of the ATTS since studies measuring 81 attitudes toward suicide by Diekstra and Kerkhof (1989) and Rogers and DeShon (1995) had been challenged to do the same. Salander-Renberg and Jacobsson decided to inspect the survey items that dealt with the referent levels of self, a close friend, or person in general, the same ones in the SUIATT. Upon comparing the referent levels of the SUIATT with those of the ATTS, Salander- Renberg and Jacobsson found the referent items to be a critical component of the survey instrument; however, a number of referent items could be modified since many of them were repetitive. In doing so, the survey items were reduced from 69 to 40 (Salander-Renberg & Jacobsson). The ATTS was then distributed to randomly selected individuals from the population (Salander-Renberg & Jacobsson, 2003). One thousand postal surveys were sent out to the general population and 64% of them were returned. Findings obtained ten factors for attitudes toward suicide: suicide as a right, incomprehensibility, non-communication, preventability, tabooing, normal- common, suicidal process, relation/caused, preparedness to prevent, and resignation. This ten factor model accounted for 60% of the total variance with internal consistency rating varying from 0.86 to 0.38. Higher variance and internal consistency rates were found for attitude factor items categorized as suicide as a right, incomprehensibility, and non-communication. According to Salander- Renberg and Jacobsson, …items from the original permissive factor model were spread out into three separate factors: suicide as a right, tabooing, and resignation. The 82 predictability and communication factors were combined into one communication factor, and the incomprehensibility factor was divided in two: incomprehensibility and the normal-common factor. The internal consistency was high (>0.70) for the two first factors and had increased for the whole instrument (p. 56). However, the internal consistency as a whole was low. In hindsight, the authors believed the goal to create a simple instrument that appealed to a broad spectrum of attitudes toward suicide that was not time consuming probably had a negative impact on the overall internal consistency of the ATTS (Salander- Renberg & Jacobsson, 2003). In effect, this instrument demonstrated an overall low rate of internal consistency which might impact the overall strength of the instrument’s reliability. In addition to the aforementioned methodological findings noted in the construction and application of the ATTS, other discussion items relating to the reliability and validity of this instrument were noted by Salander-Renberg and Jacobsson (2003). For example, the authors questioned the ability to design an instrument that measured attitudes toward suicide with high internal consistency because it was common for attitudes toward suicide to be “conflicting and unstable” (Salander-Renberg & Joacobsson, p. 61). Additonally, the ATTS was distributed to a heterogeneous population not a homogenous sample. On the other hand, the SOQ and the SUIATT were usually distributed to a homogenous sample, mostly students. Consequently, measuring the various domains 83 generated by a heterogeneous population makes it challenging to obtain results with high internal consistency. Next, the investigators concentrated on the validity of the ATTS. The 1986 and the 1996 versions of the ATTS have the same factor model structure which lends support to the strength of the construct validity. In addition, the authors sought advice from authorities in the study of suicide and the average person on the construction of the ATTS in order to assure high face validity (Salander-Renberg & Jacobsson). The reliability and validity measures were found acceptable for factors: right to commit suicide (permissiveness), incomprehensibility, non- communication, and preventability in the 1986 and the 1996 versions of the ATTS (Salander-Renberg & Jacobsson, 2003). Even though Salander-Renberg and Jacobsson reported low internal consistency measures for this instrument as a whole, the high reliability and validity measures for a handful of the factors should not be overlooked. The wide range of reliability and validity measures lends support to the belief that measuring attitudes toward suicide can be difficult because individual’s attitudes toward suicide are somewhat unpredictable or inconsistent (Salander-Renberg & Jacobsson). Finally, Salander-Renberg et al. (2008) replicated the 1986 and 1996 studies of Salander-Renberg and Jacobsson in their study, Building Models for the Relationship between Attitudes toward Suicide and Suicidal Behavior: Based on Data from General Population Surveys in Sweden, Norway, and Russia. Using the ATTS, their goal was to study the relationship between attitudes 84 toward suicide and suicidal behavior. One objective of this research is measuring attitudes toward suicide, so a review of the methodology, results, and discussions in Salander-Renberg et al. which focus on the data related to attitudes toward suicide follows. The methodology of this research was similar to previous efforts by Salander-Renberg and others discussed in this review of the literature. Salander- Renberg et al. (2008) distributed 1,000 postal surveys to individuals in Sweden and Norway, and 560 surveys to persons in Russia. Participants were residents of the general population in a county from Sweden, Norway, or Russia. All three countries used in this study were members of the “Assembly of European Regions network on suicide prevention and suicide research” (Salander-Renberg et al., p. 663). Survey response rate was as follows: (a) Sweden, 64%; (b) Norway, 44%; and (c) Russia, 92%. An assessment of the groups’ demographics was conducted by comparing background information, the samples, and the respondents from the various counties. This assessment yielded a final sample of 627 participants from Sweden, 466 participants from Norway, and 508 participants from Russia (Salander-Renberg et al.). Results of this research identified common attitude factors by using “explorative factor analysis with principal component extraction and varimax rotation” (Salander-Renberg et al., 2008, p. 666). This method generated a common six-factor model from the items on the ATTS: (a) Acceptance of suicide, (b) condemnation of suicide, (c) preventability, (d) preparedness to help, (e) 85 unpredictability, and (f) relation caused. Further, a “confirmative factor analysis” was conducted on this factor model “explaining 61% of the total variance and with a Cronbach’s alpha of 0.61 (Salander-Renberg et al., p. 667). However, the measurements of reliability for the six factors varied from very high at 0.84 to relatively low at 0.35 (Salander-Renberg et al., 2008). A similar pattern of reliability was discussed earlier in Salander-Renberg’s 1986 and 1996 studies. To address the issue of reliability, the authors used the factors with satisfactory Cronbach’s alpha measures: (a) acceptance of suicide, 0.84; (b) condemnation of suicide, 0.71; and (c) preventability of suicide, 0.64 as the standard factors used to compare individuals’ attitudes toward suicide and their suicidal behavior (Salander-Renberg et al.). To that end, the attitude factor models for the 1986 and 1996 studies conducted by Salander-Renberg and Jacobsson with the strongest reliability and validity measurements were the same ones identified in research by Salander-Renberg et al. More specifically, in all three studies, the factor known as “acceptance of suicide/suicide as a right/permissiveness” had the highest reliability and validity scores. The definitions of the attitude factors with the strongest reliability and validity scores, including suitable Cronbach’s alpha scores, are provided. These three attitude factors will be used as the independent variables in this researcher’s study. Acceptance of suicide is the belief that suicide is tolerable in situations where no other alternative is available. For example, questions from the ATTS 86 that fit this definition include, “I would consider the possibility of taking my life if I were to suffer from a severe, incurable disease,” and “There may be situations where the only reasonable resolution is suicide” (Salander-Renberg et al., 2008, p. 666). Condemnation of suicide is the exact opposite of acceptance of suicide. These individuals believe suicide is an unacceptable choice for solving one’s problems. For example, questions from the ATTS that fit this definition include, “Suicide can never be justified,” and “Committing suicide is among the worst things to do to one’s relatives” (Salander-Renberg et al., 2008, p. 666). Prevention of suicide is the belief that individuals are responsible for trying to preclude suicide in themselves and others. For example, a question on the ATTS that fits this definition is “It is always possible to help a person having suicidal thoughts” (Salander-Renberg et al., 2008, p. 666). Overall these three attitude factors look as if they demonstrate three overarching but very different positions of suicide. Individuals seem to accept suicide, condemn suicide, or try to prevent suicide. The definitions of these three attitude factors toward suicide will be used throughout this research. Next, the discussions of these results presented by Salander-Renberg et al. (2008) bear several interesting findings related to factors of acceptance of suicide, condemnation of suicide, and prevention of suicide. Relevant conclusions by the researchers stated, …age influenced attitudes, with mainly more accepting…attitudes among 87 the young. A general pattern, valid for all regions and both genders…was that older people to a lower extent agreed that it should be possible to get help to end one’s life if suffering from a severe incurable disease…Older people, those often mentioned as target groups for euthanasia , are apparently not as convinced as younger people are! Older people were also found to be more pro-preventative than young people, believing that suicide can and should be prevented. The exception was Russian women, where the young were more pro-preventative (Salander-Renberg et al., p. 672). One might conclude that it is difficult to predict individuals’ attitudes toward suicide. The wide range of reliability scores is a testament to the idea that attempting to predict who will accept suicide versus those that will condemn suicide versus those that believe suicide is preventable is difficult to predict. Finally, a comparison of specific demographic information and the attitude factors of the ATTS were analyzed in a study by Mofidi, Ghazinour, Salander- Renberg, and Richter (2008). In this cross-sectional study, the authors administered the ATTS questionnaire to 1,000 Kurdish people in Iran. Interest in finding a relationship between behavioral factors and attitudes toward suicide in order to direct suicide prevention efforts was the overall purpose of this study (Mofidi et al., 2008). A report of the analysis of demographic information in relation to attitudes toward suicide is reviewed here as it gives support to the 88 demographic information chosen to be used in conjunction with the administration of the ATTS in this study. Mofidi et al. (2008) found an association between attitude toward suicide factors and demographic information related to age, gender, marital status, and level of education. The attitude factors used in this study were acceptance of suicide, preventability of suicide, and suicide-related communication problems. Factor, suicide-related communication problems is the overarching factor category that includes condemnation of suicide (Mofidi et al.). These factors were the same ones used in previous studies by Salander-Renberg et al. (2008) and Salander-Renberg and Jacobsson (2003) discussed in this review of the literature which lends further support to their use for this study. Specific demographic information related to age, gender, marital status, and level of education was found to be significant when compared with responses to attitudes toward suicide (Mofidi et al., 2008). First, Mofidi et al. noted, “There was a significant correlation…between age and attitudes toward suicide prevention implying a rather negative…attitude toward suicide prevention with increasing age (r = 0.13; p = 0.001)” (p. 295). Second, Gender and marital status related to the factor, preventability of suicide, saw significant difference (t = 3.55; p < 0.001). Married Kurdish women reported significantly higher scores on factor, preventability of suicide when compared with their male counterparts. Third, level of education and acceptance of suicide were found to be significant (F = 8.64; p < 0.001). Individuals with an elementary school education and/or 89 senior high school education reported accepting suicide more than individuals with a college education. Furthermore, people with a college education reported higher scores (F = 7.21; p < 0.001) on the factor, difficulties communicating about suicide than their lower educated peers (Mofidi et al.). Interpretations of the Mofidi et al. (2008) study appear relevant. With respect to age of individual, and attitude factor – prevention of suicide, it seems that as an individual gets older they believe they acquire a more pessimistic attitude toward preventing suicide. This finding is in opposition to what Salander- Renberg et al. (2008) observed. Their study discovered that age had a positive impact on their attitude toward the preventability of suicide. Next, individuals with post secondary education found it difficult to talk about suicide. Mofidi et al. believed this finding might be connected to the conflicts between the societal norms of an Islamic society and the societal norms of a university community. Mofidi et al. states, “This might cause inner conflicts in the students and more highly educated people between the official societal norms and values, their expectations of their own life, and anticipated means of attaining these goals in the face of the restrictions of a conservative Islamic society” (p. 296). Age, gender, marital status, and level of education were found to be significant when compared with attitude factors of acceptance of suicide, prevention of suicide, and suicide-related communication problems (Mofidi et al. (2008). Individuals with a university education were less accepting of suicide and found it more difficult to communicate about suicide when compared to their 90 lower educated peers. Although disputed in the literature, Mofidi et al. discovered that as people age, they appear to have a more negative attitude toward preventing suicide. As such, Salander-Renberg et al. (2008), Salander-Renberg and Jacobsson (2003), and Mofidi et al. all found the strongest reliability and validity scores in attitude factors of the ATTS to be: acceptance of suicide, prevention of suicide, and suicide-related communication problems or condemnation of suicide. Furthermore, discoveries by Mofidi et al. found significance between these attitude factors and specific demographic information like age, gender, marital status, and level of education. Social Power Given the overall ill-defined ad hoc role of DOS and their role in suicide prevention on campus, an inquiry into power and leadership in this arena seems worthwhile. Effective leadership requires university administrators to be able to persuade their colleagues, stakeholders, and students (Yukl & Falbe, 1991), and the use of social power and influence has been said to accomplish this task (Raven, 1993). Historical evidence on the study of social power was well documented by Lewin (1951), Cartwright (1959), French and Raven (1959), and others. The study of social power is vast; therefore, an overview of the components of power as it is related to this study will be provided. Accordingly, this section is composed of the following information on social power: (a) the advent of social power; (b) the definitions of social power and social influence; (c) 91 defining the bases of social power; and (d) an examination of social power bases. Advent of social power. Raven (1993) acknowledges that the study of social power was initiated in research conducted by Kurt Lewin. Lewin’s studies focused on what he identified as the “power field”, specifically the dynamics of power in the relationship between the parent and the child (Raven, 1993). Lewin (1951) believed power had the potential to generate positive and negative energies of various sizes onto another person. Accordingly, these series of energies displaced on the child by the parent could potentially impose conflict on the child across his or her life span (Lewin, 1951). Building on his research of the parent-child relationship, Lewin investigated the power a group possessed over a single individual and conveyed two relevant findings (Raven, 1993). First, the collective thoughts and beliefs of the group generally contradicted the thoughts and beliefs of the individual attempting to enact change in the behavior of the group. Individuals seeking to influence the group had to approach the group as a whole in order to enact change in the group’s behavior. Second, power to convince a group to change behaviors was achieved by the individual when the individual offered the group ideas for change, guided the group in discussing why change was needed, and allowed the group to reach change on its own time frame with support from the individual seeking change (Raven, 1993). 92 Expanding on Lewin’s research, Coch and French (1968) analyzed the power individuals used to prevail over the group’s resistance to change. To test their ideas they used the employees at the Harwood Manufacturing Corporation as the subjects of their research. Their findings supported the research conclusions published by Lewin. According to Coch and French, management saw the employees resist change when management did not solicit full participation from the employees in the decision to enact change. A lack of full participation in the decision making process for change yielded a “hostile power field”. What ensued was a dictatorial management environment where the employees had no voice and felt unsatisfied by their work environment. Not surprisingly, Coch and French found that employees preferred a collaborative decision making operation, instead of a non-participatory decision making process, with management. The University of Michigan’s Research Center for Group Dynamics continued to develop research on the influence of social power on group dynamics in the 1950’s (Raven, 1993). Festinger (1954) analyzed power and the effect of sameness on groups. Lewin, Lippitt, and White (1939) initially considered the implications of power on leadership styles, but took it one step further in 1952 by investigating the characteristics summer camp participants used to influence or demonstrate power over other campers (Lippitt, Polansky, Redl, & Rosen, 1952). Hurwitz, Zander, and Hymovitch (1953) and Zander and Cohen (1955) considered the way individuals with less power appeared to think 93 suspiciously of individuals with more power. French (1956) “joined Dorwin Cartwright in applying graph and matrix theory to the analysis of group structure, an approach that Jack French utilized quite effectively in his first analysis of power structure” (Raven, 1993, pp. 231-232). It was from these research endeavors social power and interpersonal and group influence emerged as the central topic of study at the University of Michigan’s Research Center for Group Dynamics (Raven, 1993). Dorwin Cartwright set the path of research in social power in motion with his address entitled, “Power: A Neglected Variable in Social Psychology” (Cartwright, 1959). Following this publication, French and Raven conceptualized the bases of social power (Raven, 1993). The study of social power started with notable researchers like Kurt Lewin, Leon Festinger, and Dorwin Cartwright, and continued to expand after studies published by the members of the University of Michigan’s Research Center for Group Dynamics (Elias, 2008). Prior to leading up to the development of the social power bases, much of the literature focused on the study of social power and social influence. It was not until rounds of repeated discussions and research conducted by French and Raven that the bases of social power emerged (Elias). Definition of social power. French and Raven chose to examine the core dynamics of social power and social influence in order to build on the extensive research of Lewin (1951) and Festinger (1954). Ultimately, French and Raven aspired to find what means an individual uses to implement influence over 94 another (Raven, 1993). An overview of the characterization of social power and social influence is provided. French (1956) posited the initial theory of social power. French set out to discover “the extent to which the influence process in groups could be explained in terms of patterns of interpersonal relations” (p. 181). He found that the increase in the base of interpersonal power was a result of the continuing relationship between the two individuals. For the purposes of this discussion, the term “influencing agent” is the person attempting to change the “target individual” behavior, the term for the person the influencing agent is pursuing. Power varied in strength, which French identified as five bases of power. The initial five bases of power were: attraction power, expert power, reward power, coercive power, and legitimate power (French). The subsequent section of this literature review provides a description of the revised bases of social power. The application of the initial bases of social power identified by French (1956) considered three assumptions. First, in situations where opinions differed, the force one individual placed on the other individual was proportional to the strength of the base of power being used. Second, the size of the difference in opinion between the two individuals was proportional to the amount of force one individual placed on the other individual. Third, in a group situation the individual receiving the force from another individual will change his opinion after reaching “the equilibrium point” where the force by the other members and the “resisting corresponding to his own resistance to change is equal to zero” (French, p. 185). 95 Following the establishment of the original bases of social power, French and Raven (1959, 1968) investigated the outcomes of legitimate power in relationship to the outcomes of coercive power. It appeared legitimate power and coercive power both delivered change as a result of the power exerted by the influencing agent. The target individual conformed to the behavior because he or she accepted the influencing agent’s authority or he or she desired to avoid punishment. As a result, French and Raven (1959, 1968) believed coercive power required observation on the part of the influencing agent in order to change the behavior of the target individual while legitimate power did not. Leary (1957) noted that social power and social influence were inherent to human relationships. Social power was initially defined in the literature as the competence an individual possesses to wield influence over another individual’s or group’s behaviors (French & Raven, 1959, 1968). Bass (1990) defined social power as an influencing agent’s capability to cause the target individual to act in agreement with the influencing agent’s requests; this definition aligns with the one originally published by French and Raven (1959, 1968). Building on these definitions of social power, Biggert and Hamilton (1984) and Pfeffer (1981) were credited with identifying the two sources of power that exist in organizations. Individuals in an organization appear to possess a level of power due to their position or role within the organization, and their personal power which is the result of the uniqueness of their individuality. 96 Definition of social influence. In addition to social power, social influence also plays an important role in human relationships (Leary, 1957). French and Raven define social influence as the change in behavior of an individual because of another individual’s persuasion. Social dependence and the value of the surveillance appears to impact the type of influence used (Raven, 1993). For example, a supervisor who serves as the influencing agent used expertise to convince the employees, the target individuals, to change their behavior, and in doing so, the supervisor observed an increase in employee productivity (Raven, 1993). However, Raven (1993) postulates that some measures of social influence did not always correspond with the bases of social power identified initially by French (1956). For example, force, which is defined by Wrong (1980) as change that happens to an individual target without the consent of the target individual, was an outlier. The use of physical restraints without the approval of the individual target is an illustration of force (Wrong). Raven (1993) is clear to point out that the target individual’s freedom to choose to comply with the influencing agent was the distinguishing feature between using force versus using coercion. Another social influence identified by French (1956) is manipulation. This form of influence concentrates on altering the target individual’s environment in order to change the target individual’s behavior (Raven, 1993). Cartwright (1965) dubbed the ecological manipulation as an example where something in the environment was changed in order to stop the target individual’s behavior. The 97 key issue regarding both influences had to do with the fact that force and manipulation did not align with the bases of power composed by French because the target individual was left with no choices (Raven, 1993). Finally, Raven (1993) recognizes the power of third parties and the mode of influence as aspects of social influence. A supervisor could encourage change in an employee by appealing to a third party (Raven, 1993). Lewin (1952) and Coch and French (1968) detected the power of third party influence prior to Raven’s realization. The mode of influence is essentially the style the supervisor used to influence change (Raven, 1993). As a case in point, a doctor could use fearful language about the negative consequences of smoking in order to convince a patient to quit smoking. Incidentally, Raven (1993) deems this mode of influence to have the most bearing on the power base informational. An overview of social power and social influence found that organizations function in a state of being dependent on the social power and social influence of its leaders (Elias, 2008). Power and influence might be common place in an organization but defining them was not without challenge for Lewin (1952) and Cartwright (1959). After much debate over the elements of social power and social influence, French and Raven concluded that power and influence vary greatly; for example, a manager with the power to fire an employee might choose not to resort to that high level of power in order to influence the employee (Elias). Influencing agents have the ability through their behaviors on the job to increase or decrease their level of power (Kotter, 1985). 98 The bases of social power. Mintzberg (1983) and others who have conducted research on social power believe the research conducted by French and Raven on social power bases has remained the finest typology for analyzing individuals’ uses of authority. Raven (1965) modified the original bases of power from reward, coercive, attraction, expert, and legitimate (French & Raven, 1959, 1968) to reward, coercive, referent, expert, legitimate, and informational. Attraction power was renamed referent power and informational power was added to the original five bases to make six bases (Raven, 1965). An explanation of the six bases of social power – reward, coercive, referent, expert, legitimate and informational follows. Reward power is the application of positive reinforcement by the influencing agent towards the target individual, thus eliciting the desired behavior (Ambur, 2000; Raven, 1965). The influencing agent manages the power in this relationship (Ambur). The target individual perceives that the influencing agent has the power to arbitrate rewards and thus complies with the influencing agent’s requests in order to receive the perceived reward (Carson, Carson, Knight, & Roe, 1995). Coercive power is the imposition of punishment by the influencing agent toward the target individual with the intention of achieving the desired behavior (Ambur, 2000; Raven, 1965). The influencing agent manages the power in this relationship (Ambur). The target individual perceives that the influencing agent has the power to mediate the level of punishment (Carson et al., 1995). 99 Referent power is demonstrated when the target individual complies with his or her influencing agent’s requests because of the affinity the target individual has for the influencing agent (Ambur, 2000; Raven, 1965). The target individual displays respect and esteem for the personal attributes possessed by the influencing agent (Ambur). The function of referent power relies, to a large extent, on the target individual’s kinship with the influencing agent (Raven, 1993). The target individual identifies with and feels attracted to the influencing agent (Carson et al., 1995). Expert power is demonstrated by the influencing agent’s authority of professional knowledge base, and this professional posture infers power over the target individual (Ambur, 2000; Raven, 1965). The use of expert power relies on the command the influencing agent has on his or her base of knowledge and generally this power has distinct boundaries that encompass a specific skill base (Ambur). The target individual perceives that the influencing agent holds special or unusual knowledge or skills (Carson et al., 1995). Legitimate power is position-oriented authority; the influencing agent exercises power over the target individual because of his or her position of authority within the organization (Ambur, 2000; Hayden, 2000; Raven, 1965). Within organizational hierarchies, the influencing agent uses legitimate power in conjunction with other forms of power in order to change the behavior of the target individual (Ambur). The target individual perceives that the influencing 100 agent has the right to stipulate the behavior of the target individual (Carson et al., 1995). Informational power is the influencing agent’s aptitude to demonstrate a thorough understanding of the accessibility of information to the target individual (Ambur, 2000; Raven, 1965). The target individual consents to the requests of the influencing agent, but this act of approval by the target individual is not necessarily a form of appreciation for the influencing agent (Pierro, Cicero, & Raven, 2008). The complexity of this power base system lent itself to the development and interpretation of an additional seventh power base. Judged a variation of referent power, connectional power was identified by Ambur (2000). Connectional power is said to be driven by the influencing agent’s professional connections and support gained from others in the work environment. Connectional power illustrates the bandwagon effect because it generates relationships and support systems nurtured by the influencing agent (Ambur). No personal characteristics associated with the influencing agent occur. Instead, this power force is position-oriented and mirrored the characteristics of those individuals that came into contact with the influencing agent (Ambur; Hayden, 2000; Raven, 1993). It should be noted that the seventh power base, although worth mentioning since it is highlighted in the literature, was not be used in this study because it is not mentioned in The Power Base Inventory, the survey instrument used in this study. 101 Several conclusions can be extracted from the overview of the six bases of social power. The social power bases have changed in context and number since there initial conception by French and Raven (Elias, 2008). The first publication of the social power bases in 1959 uncovered five social power bases consisting of reward, coercive, legitimate, expert, and referent powers (Elias; French & Raven, 1959, 1968). It was not until 1965 that Raven convinced French that informational power was separated from expert power (Elias; Raven, 1965). The existence of the various power bases confirmed that the influencing agent’s skills to acquire and maintain power can be asserted in various modes upon the target individual (Hinkin & Schriesheim, 1994). Finally, this innovative taxonomy was meant to be a launching point from which additional research on social power could develop (Elias; Raven, 1993). Examining the social power bases. Contemporary studies have explored and investigated the potential of the social power bases in terms of influencing agents acquiring, persuading, and maintaining power over target individuals. An overview of the present-day studies of the social power bases is worth highlighting in order to provide support for the longevity of this power typology. Samples of research looking at and comparing the function of the individual power based are provided. For example, research by Ambur (2000) observed the influence of reward power and coercive power on target individuals. Influencing agents were found to control the power relationship with the target individual when reward power or 102 coercive power strategies were used. One exception to this power relationship had to do with the target individual displaying the desired behavior in which case the influencing agent’s power was deemed unessential. Reward and coercive power were primarily found to measure whether or not the target individual accomplished the given task rather than acting as a reward or punishment to the target individual (Ambur). Finally, maintaining power using reward or coercion required continued surveillance on the target individual in order to change the behavior of the target individual (Pierro et al., 2008). Subsequently, the influence of referent power on target individuals was studied. The use of referent power on the target individual did not require the continued surveillance of the influencing agent, nor did it require much action on the part of the influencing agent like that of reward or coercive power (Pierro et al., 2008; Raven, 1993). As an alternative, the target individual chose to take on the qualities of the influencing agent to comply with the influencing agent because the target individual admired the influencing agent (Pierro et al.). However, Ambur observed that some target individuals neglected to succumb to referent power in units within an organizational hierarchy. Exploration revealed that the function of referent power relied, to a large extent, on the target individual’s high regard for the influencing agent (Raven, 1993). This type of esteem was not necessarily present when an influencing agent uses expert power. Like referent power, the impact of expert power on the target did not require the continued surveillance nor did it require much action on 103 the part of the influencing agent (Pierro et al., 2008; Raven, 1993). According to Ambur (2000), expert power has distinct boundaries that encompass a specific area of knowledge. Thus, it was more defined than referent power (Ambur). An influencing agent’s use of referent and expert power has been tested by researchers. Hinkin and Schriesheim (1994) examined subordinate’s perceptions of their supervisor’s reward behaviors from the lens of the bases of power by French and Raven (1959, 1968). They determined that supervisors utilizing contingent reward behaviors on their subordinates were found to display high levels of expert and referent power. The use of contingent reward behaviors by supervisors appeared to have a strong positive effect on subordinate performance. On the other hand, in cases where supervisors did not reward the good behavior of their subordinates, the subordinates perceived this action as punitive consequently generating a negative perception of their supervisor. Hinkin and Schriesheim labeled this behavior as a form of coercive power. It should be noted that these findings align with past study by Williams and Podsadoff (1988). Referent power and expert power have changed as a result of the advances in industrial technology as reported by Ambur (2000). The revolution of the internet has provided individual’s easy access to common knowledge. The availability of this existing knowledge quite possibly could reduce the value of expert power. The influencing agent is no longer rendered the expert (Ambur). Hayden (2000) also acknowledged this idea in his research, although, he 104 contends that influencing agents might lose their expert power if their performance is perceived by targets as lacking. In this age of technology, the function of expert power appears to require expertise that is “narrowly focused” in a “highly specialized area” (Ambur). Hence, the narrow scope of expert power has the potential to increase the value of referent power. This shift in access to knowledge requires a second look at organizational power structures. Ambur maintains that organizational hierarchies should be flattened to allow for more interaction between supervisors and employees. An influencing agent’s use of legitimate power has also been evaluated in the literature. Like referent power and expert power, the impact of legitimate power on the target does not require continued surveillance nor does it require much action on the part of the influencing agent (Pierro et al., 2008; Raven, 1993). Additionally, coercive power and reward power are almost always tied in with legitimate power (Hayden, 2000). Hinkin and Schriesheim (1994) examined subordinate’s perceptions of their supervisor’s punishment behaviors using the lens of the bases of power by French and Raven (1959). Their results indicated that supervisor’s using contingent punishment behaviors on their subordinates were perceived by their subordinates to possess legitimate power. Furthermore, the supervisor’s use of punishment on subordinates enhanced the supervisor’s legitimate power especially when the subordinate knew his or her actions led to poor work performance. 105 Like referent power, expert power, and legitimate power the impact of informational power on the target individual does not require continued surveillance nor does it require much action on the part of the influencing agent (Pierro et al., 2008; Raven, 1993). Research conducted by Hayden (2000) concluded that informational power, referent power, and expert power utilized the influencing agent’s personal qualities, therefore, transforming the influencing agent into a leader. In addition, those who relied on positional oriented power were described as bureaucrats; rather, successful leaders hold the ability to implement a variety of power bases (Hayden). The examination of the social power bases studied the influencing agents’ abilities to acquire, persuade, and maintain power over target individuals. Four overarching themes emerged as a result of these investigations. First, the influencing agent was found to acquire, persuade, and maintain power over the target individual when using reward or coercive power (Ambur, 2000; Pierro et al., 2008). Second, referent power and expert power require little effort on the part of the influencing agent because the target individual admires the influencing agent and feel honored to comply with the influencing agent’s requests (Pierro et al.). Third, Ambur found that the use of referent or expert powers were not as effective with the advancements in technology because access to knowledge is considered easy to access by any employee; therefore, target individuals do not have to rely on the influencing agent to be the resident expert. Finally, Hayden 106 concluded that influencing agent’s utilizing multiple power bases were viewed as leaders within their organizations. Measuring the Social Power Bases Initial research on the social power bases spent a great deal of time defining and understanding the various power bases. Lewin (1952), Cartwright (1959), and French and Raven (1959) were credited with unraveling the characteristics of the social power bases which opened the door for research on assessing the application of the social power bases. One focus of this research was to design a measurement instrument that generated data on individuals’ use of power sources with satisfactory reliability and validity scores (Rahim, 1988). As such, research on assessing the application of the social power bases has evolved. In this section, an overview of the evolution of assessing the application of the social power bases will be presented. Aspects of the early instruments used to measure power, the development of the Leader Power Inventory, and the advance of the Power Base Inventory will be conferred. Original instruments used to measure power. Shortly after the publication of the five bases of power (French & Raven, 1959), Bachman, Smith, and Slesinger (1966) and Student (1968) designed the first research instruments used to measure the five power bases. Flaws associated with these instruments were found in its composition and findings related to validity and reliability. For example, the psychometric adequacy of these instruments was lacking because of their single item design (Rahim, 1988). As a result, little or no convergent 107 validity existed and the content validity was deemed poor because of the limitations associated with a single sample design. Finally, the overall reliability coefficients were reported to be below the acceptable level of 0.50. These instruments were single-item surveys so no internal consistency reliability could be determined, therefore, evaluating for the quality of this instrument was not possible (Rahim). These flaws in the validity and reliability of single-item instruments led researchers to consider designing multi-item instruments in an attempt to improve the validity and reliability scores for the purposes of producing a quality instrument that attempts to measure power (Nunnally, 1978). Rahim leader power inventory. Following the labors of Bachman et al. (1966) and Student (1968), Rahim (1988) designed the next instrument used to measure the power bases known as the Rahim Leader Power Inventory (RLPI). The objective of this inventory was to create a multi-item instrument that includes factors with independent subscales for measuring French and Raven’s (1959, 1968) bases of power. By designing a multi-item instrument, Rahim believed he would remedy the validity and reliability limitations of the earlier instruments. After conducting extensive trials and factor analyses to test for construct and criteria related validity and social desirability, the final version of the RLPI included 35 items which amounted to 7 items for each power base (Rahim, 1988). Using a five point Likert scale, the RLPI attempted to measure the strength of the leaders’ power base from the perception of the subordinate. High 108 scores for a particular power base were interpreted to mean the subordinate perceived the leader to be strong for that particular power source (Rahim). Distribution of the RLPI was conducted using a national sample and a collegiate sample to test the usefulness of this instrument and assess for acceptable reliability and validity levels (Rahim, 1988). First, a national sample of Directors and Executives from the Standard and Poor’s Register of Corporations were invited to answer the RLPI. Rahim obtained a response rate of 32% during two trials, n = 350 and n = 126. Results indicated “no significant non-response bias” based on tests of analyses of variance and chi-square which found no significant difference at the 0.50 level (Rahim, p. 496). In addition, a factor analysis yielded inter-correlations among the five power bases which ranged from 0.12 to 0.58 (Rahim). These inter-correlation results appear dependable as they support those found in research by Warren (1968) and Ivancevich (1970). Second, a collegiate sample was used to obtain data on the reliability, validity, and the level of social desirability responses (SDS) of the RLPI (Rahim, 1988). Undergraduate students employed with their current supervisor for a period of one year were eligible to participate in this study. A response rate of 95% (n = 297) was obtained. Results yielded reliability coefficients between 0.77 and 0.91. Furthermore, results of the Pearson’s correlations between the SDS and the RPLI revealed no social desirability response bias. These findings established that the RPLI’s reliability is satisfactory and ample evidence of criterion related validity exists. Overall the repeated factor analyses and 109 correlations conducted on the national sample and the collegiate sample demonstrate significant support for apt construct validity (Rahim). Rahim concluded that the RLPI addressed the limitations found in the original instruments used to measure the power bases. Data on reliability, validity, and freedom from social desirability response bias were improved with the design of this multi-item instrument as opposed to the original single-item instruments. According to Rahim (1988) future research using the RLPI should investigate the similarities and differences in the responses of management and non-management. Power base inventory. Building on the power base research conducted by Rahim (1988), Thomas and Thomas (1985, 1991) published the Power Base Inventory or PBI (see Appendix B). The purpose of the PBI is to assess the techniques one uses in order to influence others. One significant difference between the RLPI and the PBI exists regarding who possesses power in the organization. The RLPI assumed that power was possessed and used only by the superior; therefore, it measured a subordinate’s perceptions of his/her superior’s power. The PBI assumes that power is possessed by any member of an organization, not solely by the superior. In other words, any individual in the organization has the potential to exercise leadership or influence over another individual. It is for this reason Thomas and Thomas (1985, 1991) believe the PBI was relevant in multiple, diverse organizational settings. 110 An appraisal of the PBI was conducted by Elliot (2004) in the Mental Measurements Yearbook. Content of the PBI is fairly straightforward. A self administered, forced-choice, 30-item questionnaire takes approximately 15 minutes to complete by a group or individual who are team members, leaders, managers, or supervisors. According to Elliot, “Item statements describe why members of a team or colleagues might be influenced by the rater. Raters choose between two force-choice statements (A) and (B)” (p. 1). Six categories of power bases are measured; they are: information, expertise, goodwill (referent), authority (legitimate), reward, and discipline (coercive). Thomas and Thomas (1985, 1991) noted, “The names of some of the power bases have been changed in this inventory to better describe the form which they take in the manager’s relationship with subordinates” (p. 10). Elliot (2004) also critiqued the scoring of the PBI. Specifically, the inventory scores identify which power base the respondent is most likely to use in a given situation. Each statement (A) or (B) corresponds with a one of the six categories of power bases. The numbers of items circled in each category are summed and charted by percentile. Tallies of individuals who already took the PBI created the base of the instrument’s percentile ranks and these ranks are “divided by upper and lower 25% and middle 50%” (Elliot, p. 2). Percentile scores outside the middle 50% were not necessarily considered of concern because the use of a power base at one of the extremes could be deemed appropriate (Elliot). 111 Next, Elliot (2004) addressed the reliability and validity of the PBI. No formal analysis or explanation of reliability and validity exists. Elliot attributes this to the fact that the PBI is a “self-awareness tool that could be situational based” (p. 2). Therefore, profile scores are more often than not based on the respondents’ present state of affairs rather than a reliable predictor of consistent use of the power base. As a result profile scores can vary depending on the circumstances of the respondent. Elliot cautions users of this instrument to be attentive to this possible swing in profile scores. Finally, the PBI concludes with the framework from which one might interpret respondents’ profile scores (Thomas & Thomas, 1985, 1991). Thomas and Thomas discussed the language associated with power and leadership. The authors thought it relevant to define leadership, influence, and power within the context of the PBI. Accordingly, Thomas and Thomas defined leadership as “influencing team members to do the right thing”; influence as “altering people’s beliefs or behaviors”; and power as “the ability to influence” (p. 10). Based on these definitions Thomas and Thomas believed power was vital to successful leadership. Furthermore, power takes place in a variety of forms, some negative and some positive. In addition to building on these definitions for the purposes of interpreting the profile scores for the power bases, Thomas and Thomas identified specific uses of the power bases, warning signals associated with profile scores that are too high or to low for each power base, and requirements for increasing the use of each power base. It was the ultimate goal of the authors 112 to use the PBI to identify which form of power an individual chooses and provide suggestions for interpreting the profile scores for each power base (Thomas & Thomas). The overall findings regarding the development of instruments used to measure power have yielded several findings worth noting. Early instruments designed to measure power were thought to be weak due to their single item design. Consequently, outcomes produced disappointing reliability and validity scores (Rahim, 1988). Following initial efforts to measure power, Rahim was credited with creating the RLPI. Rahim addressed some of the problems with past power inventories by designing this multi-item instrument which included factors with independent subscales for assessing the power bases. Data from the RLPI indicated that reliability, validity, and freedom from social desirability response bias were improved (Rahim). Thomas and Thomas (1985, 1991) refined the measurement of power with their creation of the PBI. The PBI, unlike the RLPI, posited that power could be obtained by any member of an organization, including supervisors and subordinates. Thus, the PBI is an appropriate instrument to utilize in multiple, diverse organizational settings (Thomas & Thomas). Elliot (2004) indicated that because the PBI is a self-awareness tool, information on the reliability and validity of this instrument was not obtainable. Summary Measuring attitudes toward suicide is at best a challenge due to the 113 complexities or taboo associated with the act of committing suicide. The research indicates that individuals’ attitudes toward suicide tend to lean towards: (a) acceptance of suicide, (b) condemnation of suicide, and (c) ability to prevent suicide (Salander-Renberg et al., 2008; Salander-Renberg & Jacobbson, 2003). In addition, negative attitudes about suicidal individuals seem to extend to the community in the way communities tend to disregard the prevention of suicide (Ginn et al., 1988). However, Morgan et al. (1996) found that educating individuals about suicide reduced the negative attitudes individuals had toward suicide. One might conclude that individuals’ attitudes toward suicide vary and that educating people about suicide might improve suicide prevention efforts. The use of social power and social influence is said to be an effective means for leaders to persuade their colleagues (Raven, 1993; Yukl & Falbe, 1991). The first publication of the social power bases in 1959 uncovered five social power bases consisting of reward, coercive, legitimate, and referent powers (Elias, 2008; French, 1959). In 1965, Raven convinced French that informational power should be its own power separate from expert power (Elias; Raven, 1965). Much of literature examined the social powers from the view of the influencing agents’ abilities to acquire, persuade, and maintain power over their targets (Ambur, 2000; Pierro et al., 2008). The RLPI (Rahim, 1988) and the PBI (Thomas & Thomas, 1985, 1991) attempt to measure the social power bases. The construction of these multi-item instruments has yielded improvements in reliability, validity, and social desirability response bias scores. Unlike the RLPI, 114 the PBI assumes power is possessed by any member of an organization, not solely by the organization’s leader. For this reason, the PBI is appropriate to use in multiple, diverse organizational settings (Thomas & Thomas, 1985, 1991). Conclusion At their respective institutions, DOS are responsible for the development of the whole student which includes monitoring the safety and well-being of students (Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). One part of this responsibility requires DOS to pay attention to the means by which they attempt to identify students at risk for committing suicide. However, researchers have found that efforts to identify students at risk for committing suicide fall short and are deficient in producing empirical evidence regarding their effectiveness (Brener et al., 1999; Gutierrez et al., 2000; Kisch et al. 2005; Langhinrichsen-Rohling et al., 2004; Suicide Prevention Resource Center, 2004). It appears DOS’ sources of power and attitudes toward suicide affect the methods used to identify students who might commit suicide. At present, DOS’ perceived power and attitudes toward suicide are unknown. For these reasons, a better understanding of the associations between the deans’ perceived bases of power and attitudes toward suicide would appear to be valuable information for DOS in their attempt to improve the methods they use to identify students who might commit suicide. CHAPTER 3: METHODOLOGY This chapter outlines the methodology applied to understanding the association between attitudes toward suicide and bases of power in DOS. Topics in this chapter are: (a) a summary of the research questions; (b) a description of the population; (c) an explanation of the survey instruments; (d) an overview of research procedures and implementation; (e) an evaluation of the null hypotheses; and (f) an account of the data analysis. Data obtained from two modified surveys, the Attitudes toward Suicide Questionnaire (ATTS) and the Power Base Inventory (PBI) from a sample of DOS, who are members of Student Affairs Administrators in Higher Education (NASPA), were used to study the association between attitudes toward suicide and bases of power in DOS. Research Question The DOS’ position of authority and attitudes toward suicide will likely impact the methods of suicide prevention used at a given institution (Blue, 1972; Eddy et al., 1988; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). It is unknown if an association between DOS perceived attitudes toward suicide and perceived base of power exists. The overarching research question for this study is: What is the association between attitudes toward suicide and bases of power in DOS? Specifically, this study identified DOS’ attitudes toward suicide in using the three factor attitudes toward suicide model (Salander-Renberg et al., 2008) and the bases of power (French & Raven, 1968). 116 Finally, this study explored the association between gender and attitudes toward suicide and gender and the bases of power in DOS. Population The target population for this study was individuals with the employment title, DOS, who are members of the National Association for Student Personnel Administrators (NASPA) in higher education. NASPA is the international professional organization for student affairs leaders in higher education (NASPA, n.d.). Membership in NASPA includes approximately 11,000 student affairs professionals. One thousand two-hundred NASPA members are voting delegates; these delegates are senior level administrators employed in positions like vice president and/or DOS at their respective institutions (K. Kruger, personal communication, October 9, 2009). Membership types include professional affiliates, associate affiliates, faculty affiliates, graduate student affiliates, undergraduate affiliates, and emeritus affiliates who represent 2-year and 4-year institutions in the United States and abroad (NASPA, n.d.). Approximately 400 members of NASPA report having the employment title, DOS, and approximately 50 members report multiple employment titles with one being DOS (K. Kruger, personal communication, October 9, 2009). The numbers of professionals with the working title, Dean of Students, varies due to the inconsistencies in the application of this title within the national university system. NASPA members with the sole title DOS and those with the title DOS plus other employment titles were invited to participate in this study. Additional 117 characteristics of the DOS provided by NASPA include region of membership, gender, race/ethnicity, number of years in the field, and highest educational degree earned. By virtue of their membership in this professional organization and their title of employment, they are a convenient and available sample of professionals who represent the characteristics of the individuals to be evaluated (Kinnear & Gray, 2010). As a member of NASPA, permission to access the membership list of NASPA for research purposes was authorized. The procedure for obtaining the membership list of DOS required contacting Dr. Kevin Kruger, Associate Executive Director of NASPA at kkruger@naspa.org. The completion of the NASPA’s Center for Research, Membership List Request Form and a review of the Membership List Request Policy for Research Purposes located at www.naspa.org/research was required. The Membership List Request Form required the researcher to write contact information, the purpose of the research project, identification of the list of DOS characteristics to be studied, and the random sample size requested for the study. The completed Membership List Request Form, along with a copy of the institutional review board approval letter, was faxed to NASPA. The Membership List Request Policy for Research Purposes explains who is eligible to request membership lists and the criteria for receiving a NASPA membership list (see Appendix C). According to their policy, research using NASPA members must support the mission of NASPA, and the researcher must 118 be a member of NASPA. The membership list is free of charge. The researcher’s advisor must submit a letter stating his intention to supervise this study and confirm the verification of the institution’s Internal Review Board approval. Following the submission of the Membership List Request Form, in conjunction with the advisor’s letter of support for the study, NASPA reviewed this request (see Appendix D). Within three weeks of the request, NASPA provided the researcher with an Excel spreadsheet of the names of the DOS, their available demographic information, the name of the institution where they are employed, the institution’s mailing address, and their phone number. NASPA does not provide the electronic mail addresses of its members; therefore, the researcher looked up individuals’ electronic mail addresses in the NASPA electronic directory at www.naspa.org/membership. Electronic mail addresses were found for 387 members of NASPA with the working title DOS or some variation of that working title. An electronic mail correspondence was sent to the DOS informing them of the opportunity to participate in this research. Instrumentation This study utilized two survey instruments, the Attitudes toward Suicide questionnaire (ATTS), and the Power Base Inventory (PBI). The ATTS assesses individuals’ opinions about suicide, and the PBI evaluates the power bases favored by leaders in their role as supervisor. The researcher received permission to modify the instruments, and both surveys have been narrowed to meet the parameters of this particular research design. Data measuring the 119 associations between demographic factor, gender, and their corresponding attitudes toward suicide and bases of power was calculated. Attitudes towards Suicide Questionnaire The Attitudes towards Suicide Questionnaire (Salander-Renberg & Jacobsson, 2003) identifies individuals’ opinions about suicide such as the right to commit suicide, duties to prevent suicide and the suicidal process. The purpose of the ATTS is to measure people’s attitudes toward suicide through the use of a survey that is not too time consuming, could be completed by the general population, and could be modified as needed. Specific attitude variables measured in this research include: (a) acceptance of suicide, (b) condemnation of suicide, and (c) prevention of suicide (Salander-Renberg & Jacobsson). The ATTS (Salander-Renberg & Jacobsson, 2003) was modified from its original format of three sections constituting 61 statements. Statements 4 – 43 in the ATTS are concerned with respondents’ attitudes toward suicide. For the purposes of this research, the ATTS was reduced to nine statements. Respondents answered items 4, 5, 6, 8, 12, 19, 30, 37, and 40 as they pertained to the acceptance, condemnation, and prevention of suicide. Additional variables that measure attitudes toward suicide in the ATTS that were not used in this study include ability to predict suicide, perceptions about communicating suicide, opinions about reasons for attempting suicide, and ideas concerning the duration of the suicide process. Exploration of these factors was excluded as they are not germane to the present inquiry. 120 A Likert scale (strongly agree, agree, undecided, disagree, or strongly disagree) is used to identify the participants response to the statement about suicide. Within the survey instrument, statements 8, 19, and 37 deal with attitude variable, acceptance of suicide; statements 5, 6, and 30 deal with attitude variable, condemnation of suicide; and statements 4, 12, and 40 deal with attitude variable, prevention of suicide. These nine statements in this modified version of the ATTS serve to support the purposes of this study. Power Base Inventory The Power Base Inventory (Thomas & Thomas, 1991) identifies the perceived source of power utilized by an individual in his or her role as employee. More to the point, the PBI quantifies individuals’ abilities to influence the behavior of other team members. The purpose of this survey is to identify which of the six power bases (a) reward, (b) coercion (discipline), (c) legitimate (authority), (d) referent (goodwill), (e) expert, and (f) information the individual uses when trying to convince other team members to conform to his or her wants (Thomas & Thomas). The PBI (Thomas & Thomas, 1991) was modified from its original format of thirty pairs of statements. The thirty pairs of statements in the PBI intend to identify the respondents’ bases of power. For the purposes of this research, the PBI was reduced to fifteen pairs of statements. While the thirty pairs are used to enhance consistency, I concluded a higher participation rate was garnered 121 among the DOS population with a more streamlined instrument. Consistency was analyzed against the findings of previous PBI respondents as described below. Respondents were asked to choose the statement (A or B) that most describes the reason a team member might be influenced by the respondent; thereby, this instrument is a forced-choice survey. Upon completion of the inventory, the respondents’ scores for each power base are calculated and reported. The total score is then graphed side-by-side in relation to the findings of the 317 managers who have already taken the PBI. Respondents found to score in the middle 50% range for that power base are said to be in line with the responses of the 317 managers who already took the PBI. Outlying scores or scores that fall above or below the 50% range for that power base indicate respondents use that power base either more or less than the 317 managers who already took the PBI. Thomas and Thomas note that outlying scores should not be viewed as a flaw because individual circumstances might substantiate high or low use of a given power base. At the conclusion of the inventory the authors provide the researcher with suggestions for interpreting the power base scores including: (a) when to use the power bases, (b) what to do if power base scores are too high or too low, and (c) suggestions for how to increase the use of the power base (Thomas & Thomas, 1991). 122 Procedures and Implementation The procedure for obtaining permission to use the ATTS and PBI started in the spring of 2009. Permission to use the ATTS simply required contacting the primary author of this instrument, E. Salander-Renberg, Ph. D., Department of Clinical Sciences, Division of Psychiatry, Umea University, Sweden. A more formal process was required for obtaining permission to use the PBI through the company, Consulting Psychologists Press, Inc. (CPP, Inc.). The procedures and implementation of this study included the following: (a) the process for securing permission to utilize the research instruments, the ATTS and the PBI, (b) the procedures for gaining approval from the Internal Review Board, and (c) the method for administering the survey instruments. Procedure for obtaining research instruments. The procedure for obtaining permission to use the ATTS was initiated by the researcher through electronic mail correspondences with the primary author of this instrument. In a reply to the researcher, the author (E. Salander-Renberg, personal communication, May 5, 2009) granted the researcher permission to use the ATTS. In addition, the author provided the researcher with a copy of the ATTS in a PDF format along with the author’s latest research article that utilized the ATTS (E. Salander-Renberg, personal communication, May 5, 2009). The instructions included with this instrument grant the researcher permission to modify the instrument and utilize the section(s) of the instrument germane to this study (Salander-Renberg & Jacobsson, 2003). 123 The procedure for obtaining permission to use the PBI was initiated by the researcher through electronic mail correspondences with CPP, Inc.’s Permissions Coordinator, Ms. E. McLane. As per CCP, Inc. policy, permission to use the PBI in an electronic format required the completion of the following documents: CPP, Inc. Eligibility Form, CCP, Inc. Request for Adaptation of Materials for Electronic Delivery, and the CCP, Inc. Request for Research Support. The CPP, Inc. Eligibility Form and the CPP, Inc. Request for Adaptation of Materials for Electronic Delivery documents are short applications that ask for general information about the researcher, the title of the researcher project, and the type of inventory requested. The CCP, Inc. Request for Research Support is an abstract of the study including the project’s title, problem statement, brief literature review, researcher procedures, data analysis, time-line for completion, and nature of support needed from CCP, Inc. The abovementioned application forms/documents were completed by the researcher and submitted electronically to the CPP, Inc. Permissions Coordinator (see Appendix E). The CPP, Inc. Permissions Coordinator sent the researcher the Permission Agreement for Adaptation and Electronic Delivery contract which granted the researcher permission to use the PBI in an electronic format for this study (E. McLane, personal communication, August 25, 2009). The researcher was instructed to sign the contract and return it to the CCP, Inc. Permissions Coordinator. Then, an authorized representative from CCP signed the contract and the CCP Permissions Coordinator (E. McLane, personal communication, 124 August 31, 2009) returned the contract to the researcher (see Appendix F). This contract was amended to reflect the number of participants eligible to participate in this study prior to launching the survey (see Appendix G). Internal review board procedures. The University’s Institutional Review Board (IRB) procedures were followed to assess research risk and review of the study. A copy of the IRB letter of approval can be found in Appendix H. Survey administration. To ensure proper functioning of the online survey instrument used for this research, the instrumentation for this study was piloted in electronic format to five individuals who were not eligible to participate in this research. The electronic format of the instrumentation used in the pilot study was identical to the electronic format of the instrumentation used in the actual study. No data from the pilot study was analyzed. The pilot study confirmed the operational accuracy of the instrumentation and no modifications were required. Following the completion of the pilot study, and one week before the study began, a brief electronic correspondence was sent to the eligible participants of this study explaining the purpose of this research. Participants were instructed to consider power in association with attitudes toward suicide when completing the survey. This electronic correspondence allowed the researcher to determine the accuracy of participants’ electronic mail addresses. Electronic mail addresses that were returned were corrected and added to the spread sheet, or deleted if no correct electronic mail address was available. 125 Instrumentation for this study was formatted using the password protected electronic survey system called Perseus at www.perseus.com. An invitation to take part in this study was sent to eligible participants two weeks after the start of the spring 2010 semester. This email included: (a) an introduction of the research, containing the purpose of the study and the benefits to volunteering to participate in this study, (b) a confidentiality statement, and (c) directions for accessing the survey along with a web link to the survey (see Appendix H). Approximately three days after sending out the initial invitation to participate in this study, a second invitation was sent to the participants who had not completed the survey, encouraging them to participate in this study. Three days following the second invitation, a third and final invitation for individuals to participate in this study was sent to potential respondents who had not completed the survey. Null Hypotheses For the first 18 null hypotheses, the independent variable is one of the following power bases: (a) information, (b) expertise, (c) referent (goodwill), (d) legitimate (authority), (e) reward, and (f) coercive (discipline). For each of the 18 null hypotheses the dependent variable is one of the three factors of attitudes towards suicide: (a) acceptance, (b) condemnation, and (c) prevention. The set method of a one-sample chi-square test was applied to each of the first 18 null hypotheses (Salkind, 2004). The individuals from the sample represent the cases for these hypotheses. Each case, or individual, can have one variable, attitude 126 toward suicide, with six values, the power bases. This non-parametric test is appropriate when the researcher is attempting to find significance with diverse categorical data in frequency count format from a large sample size (Salkind). H0 1 There is no statistically significant association between the power base, information, and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, Information, and the attitude toward suicide, acceptance. The independent variable is the power base, information, and the dependent variable is the attitude toward suicide, acceptance. H0 2 There is no statistically significant association between the power base, expertise, and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, expertise, and the attitude toward suicide, acceptance. The independent variable is the power base, expertise, and the dependent variable is the attitude toward suicide, acceptance. H0 3 There is no statistically significant association between the power base, referent (goodwill), and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. 127 This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, referent (goodwill), and the attitude toward suicide, acceptance. The independent variable is the power base, referent (goodwill), and the dependent variable is the attitude toward suicide, acceptance. H0 4 There is no statistically significant association between the power base, legitimate (Authority), and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, legitimate (authority), and the attitude toward suicide, acceptance. The independent variable is the power base, legitimate (authority), and the dependent variable is the attitude toward suicide, acceptance. H0 5 There is no statistically significant association between the power base, reward, and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, reward, and the attitude toward suicide, acceptance. The independent variable is the power base, reward, and the dependent variable is the attitude toward suicide, acceptance. 128 H0 6 There is no statistically significant association between the power base, coercive (discipline), and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, coercive (discipline), and the attitude toward suicide, acceptance. The independent variable is the power base, coercive (discipline), and the dependent variable is the attitude toward suicide, acceptance. H0 7 There is no statistically significant association between the power base, information, and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, information, and the attitude toward suicide, condemnation. The independent variable is the power base, information, and the dependent variable is the attitude toward suicide, condemnation. H0 8 There is no statistically significant relationship between the power base, expertise, and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, expertise, and the attitude toward suicide, condemnation. The 129 independent variable is the power base, expertise, and the dependent variable is the attitude toward suicide, condemnation. H0 9 There is no statistically significant association between the power base, referent (goodwill), and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, referent (goodwill), and the attitude toward suicide, condemnation. The independent variable is the power base, referent (goodwill), and the dependent variable is the attitude toward suicide, condemnation. H0 10 There is no statistically significant association between the power base, legitimate (authority), and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, legitimate (authority), and the attitude toward suicide, condemnation. The independent variable is the power base, legitimate (authority), and the dependent variable is the attitude toward suicide, condemnation. H0 11 There is no statistically significant association between the power base, reward, and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. 130 This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, reward, and the attitude toward suicide, condemnation. The independent variable is the power base, reward, and the dependent variable is the attitude toward suicide, condemnation. H0 12 There is no statistically significant association between the power base, coercive (discipline), and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, coercive (discipline), and the attitude toward suicide, condemnation. The independent variable is the power base, coercive (discipline), and the dependent variable is the attitude toward suicide, condemnation. H0 13 There is no statistically significant association between the power base, information, and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, information, and the attitude toward suicide, prevention. The independent variable is the power base, information, and the dependent variable is the attitude toward suicide, prevention. 131 H0 14 There is no statistically significant association between the power base, expertise, and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, expertise, and the attitude toward suicide, prevention. The independent variable is the power base, expertise, and the dependent variable is the attitude toward suicide, prevention. H0 15 There is no statistically significant association between the power base, referent (goodwill), and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, referent (goodwill), and the attitude toward suicide, prevention. The independent variable is the power base, referent (goodwill), and the dependent variable is the attitude toward suicide, prevention. H0 16 There is no statistically significant association between the power base, legitimate (authority), and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, legitimate (authority), and the attitude toward suicide, prevention. 132 The independent variable is the power base, legitimate (authority), and the dependent variable is the attitude toward suicide, prevention. H0 17 There is no statistically significant association between the power base, reward, and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, reward, and the attitude toward suicide, prevention. The independent variable is the power base, reward, and the dependent variable is the attitude toward suicide, prevention. H0 18 There is no statistically significant association between the power base, coercive (discipline,) and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. This hypothesis was tested using the standard method one-sample chi- square test to test the null hypothesis that there is no association between the power base, coercive (discipline), and the attitude toward suicide, prevention. The independent variable is the power base, coercive (discipline), and the dependent variable is the attitude toward suicide, prevention. For null hypotheses 19 and 20 an independent-samples t-Test was used. The independent-sample t-Test is appropriate in determining significance within the average scores of the variables, power bases and attitudes toward suicide, and the two groups, male and female (Salkind, 2004). The independent variable 133 is gender and the dependent variables are the power bases and the attitudes toward suicide, respectively. Gender was used as a comparative factor in the current analyses for good reason. Studies by Mofidi et al. (2008) and Salander- Renberg et al. (2008) reported descriptive data on individuals’ attitudes toward suicide in their respective studies. Salander-Renberg et al. and Mofidi et al. found that an individuals’ gender had an impact on the preventability of suicide. Specifically, women reported significantly higher scores on factor, preventability of suicide when compared with their male counterparts. This finding provides justification for using gender as a comparative factor in the analyses of this research. Rejection of null hypotheses 19 and 20 will show that the independent variable, gender, is a statistically significant factor in determining the dependent variable, power bases and attitudes toward suicide, respectively. H0 19 There is no statistically significant difference betweem the gender and the six bases of power in Deans of Students who are members of NASPA. This hypothesis was tested using an independent-samples t-Test to test the null hypothesis that there is no association between gender (male and female) and the six bases of power. The independent variables are male and female, and the dependent variables are information, expertise, referent (goodwill), legitimate (authority), reward, and coercive (discipline). H0 20 There is no statistically significant difference in gender and attitude toward suicide in Deans of Students who are members of NASPA. 134 This hypothesis was tested using an independent-samples t-Test to test the null hypothesis that there is no association between gender (male and female) and the attitudes toward suicide. The independent variables are male and female, and the dependent variables are acceptance, condemnation, and prevention. Data Analysis Data collected from respondents were imported, reviewed, and managed by the researcher through the Perseus electronic survey system. Participants in this study provided their responses directly into the Perseus system. The researcher managed this password protected survey site in order to maintain the confidentiality and security of the respondents’ answers. After the survey was closed, data was exported from Perseus directly into the Predictive Analytics SoftWare (PASW) system for analysis. Before analyzing the data, variables were named and labeled (Kinnear & Gray, 2010). Next, variables were assigned a level of measurement. Survey statements having to do with the ATTS and PBI were assigned the level of measurement scale. Survey statements having to do with gender and highest degree earned were assigned the level of measurement nominal. Finally, the first nine statements having to do with attitudes toward suicide were transformed into vectors labeled, acceptance, condemnation and prevention (Kinnear & Gray). Participants’ responses on the PBI were sent to the research associates at CCP, Inc. for scoring through the 135 PASW. Results of the PBI scores were returned to the researcher electronically on the PASW. Participants’ responses to survey statements were summarized. The number of valid responses and corresponding means and standard deviations for each survey statement was reported. Overall response rate along with demographic data related to gender and highest degree earned was provided in detail. Data was analyzed to determine the presence of statistically significant associations between attitudes toward suicide variables (acceptance, condemnation and prevention) and bases of power variables (information, expertise, referent/goodwill, legitimate/authority, reward, and coercive/discipline. In addition, statistical significance for the association between gender and attitudes toward suicide and gender and bases of power was analyzed. Conclusion Suicide is the second leading cause of death among college students (Silverman et al., 1997; U.S. Department of Education National Center for Education Statistics, 2001). There appears to be a history of inadequate efforts made by administrators at colleges and universities to identify students at risk for committing suicide (The Jed Foundation, 2006; Meilman et al., 1994; Webb, 1986; Westefeld & Pattillo, 1987). As well, activities intended to identify students at risk for committing suicide have failed to yield universally accepted standards for dealing with students in crisis. DOS appear to be the leaders in student affairs with the power to influence the methods used to identify students at risk for 136 committing suicide (Blue, 1972; Eddy et al., 1988; Komives et al., 2003; Nuss, 2003; Sandeen, 1991; Sandeen & Barr, 2006). It is unknown whether or not efforts to identify students at risk for committing suicide are guided by the DOS opinions about suicide and position of authority in their role as an administrator in student affairs. A better understanding of the associations between the deans’ perceived bases of power and attitudes toward suicide would appear to be valuable information for DOS in their attempt to improve the methods they use to identify students who might commit suicide. CHAPTER 4: RESULTS Introduction The purpose of this study is to investigate the association between Deans of Students’ (DOS) attitudes toward suicide and their bases of power. Twenty null hypotheses were created to test the research question: What is the association between attitudes toward suicide and bases of power in DOS? The independent variables are the six bases of power, and the dependent variables are the three attitudes toward suicide. Findings of this study suggest DOS who believe suicide is preventable tend to perceive their base of power as expertise, legitimate, or reward. In addition, DOS with attitudes of condemnation toward suicide tend to perceive their base of power as expertise. This chapter consists of the following: (a) an overview of participants’ responses, (b) an analysis of the data, (c) a summary of significant findings, and (d) challenges with the instrumentation. Participants’ Responses In the spring of 2010, 385 members of NASPA with the working title, Dean of Students (or a similar variation of that title), were invited to participate in this study. Participants were asked to complete an on-line modified survey instrument using the Perseus web-based survey software system. The modified survey instrument included two demographic questions, nine items from the Attitudes toward Suicide Questionnaire (ATTS) (Salander-Renberg & Jacobsson, 2003), and fifteen items from CPP, Inc.’s Power Base Inventory (PBI) (Thomas, 1985). 138 Of the 385 individuals eligible to take part in this research, 219 individuals opened the survey and 185 people responded to the survey items. Thirty-four participants opened the survey but elected not to complete any survey items. The overall response rate based on the total responses collected (n = 185) was 48%. This response rate is considered good according to Bosnjak and Tuten (2001) and Bassi (1991) who report that a high-quality response rate to an online survey is 50%. In addition, a response rate of 48% is consistent with research conducted by Sheehan (2001) who found response rates ranged from 24% to 61% when reviewing the response rates for e-mail surveys conducted from 1986 - 2000. Demographic information of gender and highest degree earned was provided by the participants. Of the 185 total responses, 95 (51.6%) of the respondents self-reported as female and 89 (48.4%) as male. One respondent did not provide information about gender. In addition, 92 (50%) self-reported as having earned a doctorate, 84 (45.7%) as having earned a master’s, 5 (2.7%) as having earned a post master’s certificate, and 3 (1.6%) as having earned a bachelor’s. No respondents reported earning an associate’s. One respondent did not provide information about highest degree earned. Demographic information can be found in Table 1. Survey participants evaluated nine statements from the Attitudes toward Suicide Questionnaire (ATTS) (Salander-Renberg & Jacobsson, 2003). Of the nine statements from the ATTS, three opinion statements are concerned with acceptance of suicide; three statements are concerned with condemnation of 139 Table 1 Gender and Highest Degree Earned of Participants N % Gender Female 95 51.6 Male 89 48.4 No Response 1 0.01 Highest Degree Earned Doctorate 92 50 Master’s 84 45.7 Post Master’s Certificate 5 2.7 Bachelors 3 1.6 No Response 1 0.01 Total 185 100 140 suicide; and three statements are concerned with prevention of suicide. Using a 5-point Likert scale, strongly agree, agree, undecided, disagree, strongly disagree, participants provided their opinion about the statement as it related to suicide. The following is a synopsis of the participant’s responses to the nine items. The number of valid responses to acceptance of suicide statement, Suicide is an acceptable means to terminate an incurable disease, is 184 (M = 3.00, SD = 1.12). Ten (5.4%) of the respondents strongly agree, 64 (34.8%) agree, 44 (23.9%) are undecided, 45 (24.5%) disagree, and 21 (11.4%) strongly disagree with this statement. One participant did not respond to this statement. Although the majority of the respondents agree with this statement, the number of respondents who disagree with this statement is about equal to the number of respondents who are undecided about this statement. The number of valid responses to acceptance of suicide statement, There may be situations where the only reasonable resolution is suicide, is 185 (M = 3.44, SD = 108). Three (1.6%) of the respondents strongly agree, 43 (23.2%) agree, 40 (21.6%) are undecided, 66 (35.7%) disagree, and 33 (17.8%) strongly disagree with this statement. Opinions appear to be somewhat evenly distributed between agreeing and disagreeing with this statement; however, slightly more of the respondents disagree with this statement. 141 The number of valid responses to acceptance of suicide statement, People do have the right to take their own lives, is 184 (M = 2.73, SD = 1.04). Fifteen (8.2%) of the respondents strongly agree, 77 (41.8%) agree, 43 (23.4%) are undecided, 40 (21.7%) disagree, and 9 (4.9%) strongly disagree with this statement. One participant did not respond to this statement. Although the majority of the respondents agree with this statement, the number of respondents who disagree with this statement is about equal to the number of respondents who are undecided about this statement. The number of valid responses to condemnation of suicide statement, Suicide can never be justified, is 185 (M = 2.67, SD = 1.17). Thirty-five (18.9%) of the respondents strongly agree, 55 (29.7%) agree, 37 (20.0%) are undecided, 51 (27.6%) disagree, and 7 (3.8%) strongly disagree with this statement. Opinions appear to be somewhat evenly distributed between agreeing and disagreeing with this statement; however, slightly more of the respondents agree with this statement. The number of valid responses to condemnation of suicide statement, Committing suicide is among the worst things to do to one’s relatives, is 185 (M = 2.23, SD = 1.04). Forty-five (24.3%) of the respondents strongly agree, 86 (46.5%) agree, 23 (12.4%) are undecided, 27 (14.6%) disagree, and 4 (2.2%) strongly disagree with this statement. Respondents overwhelmingly agree and strongly agree with this statement. 142 The number of valid responses to condemnation of suicide statement, On the whole, I do not understand how people can take their lives, is 185 (M = 3.33, SD = 1.13). Ten (5.4%) of the respondents strongly agree, 50 (27.0%) agree, 12 (6.5%) are undecided, 95 (51.4%) disagree, and 18 (9.7%) strongly disagree with this statement. Respondents overwhelmingly disagree and strongly disagree with this statement. The number of valid responses to prevention of suicide statement, It is always possible to help a person with suicidal thoughts, is 185 (M = 2.85, SD = 1.23). Twenty-seven (14.6%) of the respondents strongly agree, 63 (34.1%) agree, 14 (7.6%) are undecided, 71 (38.4%) disagree, and 10 (5.4%) strongly disagree with this statement. Opinions appear to be somewhat evenly distributed between agreeing and disagreeing with this statement; however, slightly more of the respondents disagree with this statement. The number of valid responses to prevention of suicide statement, It is a human duty to try to stop someone from committing suicide, is 184 (M = 1.85, SD = 0.82). Sixty-three (34.2%) of the respondents strongly agree, 99 (53.8%) agree, 9 (4.9%) are undecided, 12 (6.5%) disagree, and 1 (0.5%) strongly disagree with this statement. One participant did not respond to this statement. Respondents overwhelmingly agree and strongly agree with this statement. The number of valid responses to prevention of suicide statement, Suicide can be prevented, is 185 (M = 2.1, SD = 0.09). Thirty-nine (21.1%) of the respondents strongly agree, 104 (56.2%) agree, 14 (7.6%) are undecided, 143 27 (14.6%) disagree, and 1 (0.5%) strongly disagree with this statement. Respondents overwhelmingly agree and strongly agree with this statement. In sum, respondents overwhelmingly believe that suicide can be prevented. In addition, respondents tend to believe individuals have the right to take their own life, and respondents seem to understand why individuals might take their life. Survey participants evaluated fifteen items from CPP, Inc.’s, Power Base Inventory (PBI) (Thomas, 1985). This modified format of the PBI is comprised of 15 paired, ipsative, forced-choice statements. Respondents chose statement A or statement B based on their opinion of how their subordinates perceive their authority. Each statement involved pairing the bases of power: Information, Expertise, Referent (Goodwill), Legitimate (Authority), Reward, Coercion (Discipline) with one another one time. The number of valid responses to statement, (A-expertise), They think I am much smarter about these things, or (B-information), They have to agree with the facts that I use for support, is 185 (M = 1.76, SD = 0.42). Forty-three (23.2%) of the respondents selected statement A, and 142 (76.8%) selected statement B. An overwhelming majority of the respondents believe their subordinates comply with them because their subordinates have to agree with the facts that the respondents use for support. This finding suggests that when presented with these two statements respondents reported a preference for the information power base. 144 The number of valid responses to statement, (A-legitimate), They accept my formal right to decide matters, or (B-referent), They have a general sense of goodwill toward me, is 185 (M = 1.60, SD = 0.49). Seventy-four (40.0%) of the respondents selected statement A, and 111 (60.0%) selected statement B. A majority of the respondents believe their subordinates comply with them because their subordinates have a general sense of goodwill towards the respondents. This finding suggests that when presented with these two statements respondents reported a preference for the referent power base. The number of valid responses to statement, (A-reward), They believe that I may do something for them in return for their assistance, or (B-coercive), They realize that, beyond a certain point, noncompliance might not be tolerated, is 184 (M = 1.78, SD = 0.41). Thirty-nine (21.2%) of respondents selected statement A, and 145 (78.8%) selected statement B. One respondent did not answer this item. An overwhelming majority of the respondents believe their subordinates comply with them because their subordinates realize that, beyond a certain point, noncompliance might not be tolerated. This finding suggests that when presented with these two statements respondents reported a preference for the coercive power base. The number of valid responses to statement, (A-referent), They enjoy doing what they can for me, or (B-expertise), They are impressed with my greater competence, is 183 (M = 1.16, SD = 0.37). One-hundred fifty-two (83.1%) of respondents selected statement A, and 31 (16.9%) selected statement B. Two 145 respondents did not answer this item. An overwhelming majority of the respondents believe their subordinates comply with them because their subordinates enjoy doing what they can for the respondents. This finding suggests that when presented with these two statements respondents reported a preference for the referent power base. The number of valid responses to statement, (A-legitimate), They believe my official status allows me to settle these issues, or (B-reward), They know I will try to make their cooperation worthwhile for them, is 184 (M = 1.32, SD = 0.46). One-hundred twenty-five (67.9%) of respondents selected statement A, and 59 (32.1%) selected statement B. One respondent did not answer this item. An overwhelming majority of the respondents believe their subordinates comply with them because their subordinates think their respondents’ official status allows the respondents to settle these issues. This finding suggests that when presented with these two statements respondents reported a preference for the legitimate power base. The number of valid responses to statement, (A-information), If we disagree, I demonstrate to them how they are wrong, or (B-coercive), If things got out of hand, they know I would have to be firm with them for the good of the organization, is 182 (M = 1.18, SD = 0.39). Thirty-four (18.7%) of respondents selected statement A, and 148 (81.3%) selected statement B. Three respondents did not answer this item. An overwhelming majority of the respondents believe their subordinates comply with them because if things get out of hand, the 146 subordinates know the respondents would have to be firm with them for the good of the organization. This finding suggests that when presented with these two statements respondents reported a preference for the coercive power base. The number of valid responses to statement, (A-expertise), They are impressed with my greater competence, or (B-legitimate), They feel formally responsible for following my instructions, is 178 (M = 1.85, SD = 0.35). Twenty- six (14.6%) of respondents selected statement A, and 152 (85.4%) selected statement B. Seven of the respondents did not answer this question. An overwhelming majority of the respondents believe their subordinates comply with them because their subordinates feel formally responsible for following the respondent’s instructions. This finding suggests that when presented with these two statements respondents reported a preference for the legitimate power base. The number of valid responses to statement, (A-referent), They comply because they care about me and like to make me happy, or (B-reward), They perceive that I will reward them for helping me out, is 181 (M = 1.24, SD = 0.43). One-hundred thirty-six (75.1%) of respondents selected statement A, and 45 (24.9%) selected statement B. Four respondents did not answer this item. An overwhelming majority of the respondents believe their subordinates comply with them because their subordinates care about the respondents and like to make their respondents happy. This finding suggests that when presented with these two statements respondents reported a preference for the referent power base. 147 The number of valid responses to statement, (A-coercive), They know they would have to be punished if they violated important directives, or (B-expertise), They put less stock in their own powers of judgment than in mine, is 178 (M = 1.52, SD = 0.50). Eighty-four (47.2%) of respondents selected statement A, and 94 (52.8%) selected statement B. Seven respondents did not answer this item. The participants’ valid responses are almost evenly divided between statement A and statement B. This finding suggests that when presented with these two statements respondents reported a slight preference for the expertise power base. The number of valid responses to statement, (A-information), They understand my reasoning and are persuaded by it, or (B-referent), I have rapport with them, is 184 (M = 1.59, SD = 0.49). Seventy-five (40.8%) of respondents selected statement A, and 109 (59.2%) selected statement B. One respondent did not answer this item. A majority of the respondents believe their subordinates comply with them because the respondents feel they have rapport with their subordinates. This finding suggests that when presented with these two statements respondents reported a preference for the referent power base. The number of valid responses to statement, (A-coercive), They believe that I could be hard on them if they deserved it, or (B-legitimate), They believe that it is their duty to obey me, is 181 (M = 1.70, SD = 0.45). Fifty-four (29.8%) of respondents selected statement A, and 127 (70.2%) selected statement B. Four respondents did not answer this item. An overwhelming majority of the 148 respondents believe their subordinates comply with them because it is the subordinates’ duty to obey the respondents. This finding suggests that when presented with these two statements respondents reported a preference for the legitimate power base. The number of valid responses to statement, (A-reward), They see that I provide positive incentives for their contributions, or (B-information), I show them how to properly interpret and deal with the situation, so we agree, is 179 (M = 1.69, SD = 0.46). Fifty-four (30.2%) of respondents selected statement A, and 125 (69.8%) selected statement B. Six respondents did not answer this item. An overwhelming majority of the respondents believe their subordinates comply with their requests because the respondents show the subordinates how to properly interpret and deal with the situation, so that both parties agree. This finding suggests that when presented with these two statements respondents reported a preference for the information power base. The number of valid responses to statement, (A-referent), They comply because they care about me and like to make me happy, or (B-coercive), They are aware that if they persisted in defying me, I might have to take corrective action for everyone’s sake, is 182 (M = 1.50, SD = 0.50). Ninety (49.5%) of respondents selected statement A, and 92 (50.5%) selected statement B. Three respondents did not answer this item. The participants’ valid responses are almost divided evenly between statement A and statement B. This finding 149 suggests that when presented with these two statements respondents reported a slight preference for the coercive power base. The number of valid responses to statement, (A-legitimate), They believe that my official status allows me to settle these issues, or (B-information), They have to agree with the facts that I use for support, is 183 (M = 1.30, SD = 0.46). One-hundred twenty-eight (69.9%) of respondents selected statement A, and 55 (30.1%) selected statement B. Two respondents did not answer this item. An overwhelming majority of the respondents believe their subordinates comply with their requests because the subordinates believe that the respondents’ official status allows the respondents to settle issues. This finding suggests that when presented with these two statements respondents reported a preference for the legitimate power base. The number of valid responses to statement, (A-expertise), They trust my skills and abilities much more than their own, or (B-reward), They realize that conscientious cooperation will merit some form of compensation, is 177 (M = 1.32, SD = 0.47). One-hundred nineteen (67.2%) of respondents selected statement A, and 58 (32.8%) selected statement B. Eight respondents did not answer this item. This item reported the lowest number of valid responses. An overwhelming majority of the respondents believe their subordinates comply with their requests because the subordinates trust the respondents’ abilities much more than their own. This finding suggests that when presented with these two statements respondents reported a preference for the expertise power base. 150 When scored as a group, results of the PBI indicate that respondents strongly prefer referent and legitimate power. Specifically, graphing total scores in relation to the scores of managers who have already taken the PBI yielded the preferred power bases, referent and legitimate at a rate of 80%. This figure indicates that respondents scored higher than 80% of the individuals who have taken the PBI; therefore, the respondents in this study are in the top 20% of participants who prefer to use power bases, referent and legitimate (Thomas, 1985). Analysis of Data The Predictive Analytics SoftWare package (PASW) 17.0 was used to analyze the data. Prior to running tests for significance, three vectors were created by the researcher to transform and condense the data from the nine statements from the ATTS to three variables, acceptance, condemnation, and prevention. The research associates at CPP, Inc. scored the PBI and provided the researcher with power base scores for variables, information, expertise, referent (goodwill), legitimate (authority), reward, and coercive (discipline). A one-sample chi-square test was applied to each of the first 18 null hypotheses to determine if there was a statistically significant association between the independent variables (bases of power) and the dependent variables (attitudes toward suicide). Crosstab analyses of these hypotheses using the Pearson Chi-Square analyzed the association between the bases power and the attitudes toward suicide when n = 185. 151 Null Hypothesis One There is no statistically significant association between the power base, information, and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of acceptance used the base of power, information. The chi-square scores of the information group, ?² (3, N = 185) = 99.43 present no association with the scores of the acceptance group, ?² (12, N = 185) = 91.65. Further, a crosstab analysis yields no statistical significance, ?² (36, N = 185) = 33.58, p = 0.58. Results of the Pearson Chi-Square analysis indicate no statistically significant association between variables, information and acceptance; therefore, H0 1 was not rejected. Null Hypothesis Two There is no statistically significant association between the power base, expertise, and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of acceptance used the base of power, expertise. The chi-square scores of the expertise group, ?² (5, N = 185) = 119.70 present no association with the scores of the acceptance group, ?² (12, N = 185) = 91.65. Further, a crosstab analysis yields no statistical significance, ?² (60, N = 185) = 66.63, p = 0.25. Results of the Pearson Chi-Square analysis indicate no statistically significant associations 152 between variables, expertise and acceptance; therefore, H0 2 was not rejected. Null Hypothesis Three There is no statistically significant association between the power base, referent (goodwill,) and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of acceptance used the base of power, referent (goodwill). The chi-square scores of the referent (goodwill) group, ?² (5, N = 185) = 112.76 present no association with the scores of the acceptance group, ?² (12, N = 185) = 91.65. Further, a crosstab analysis to determine association yields no statistical significance, ?² (60, N = 185) = 50.86, p = 0.79. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, referent (goodwill) and acceptance; therefore, H0 3 was not rejected. Null Hypothesis Four There is no statistically significant association between the power base, legitimate (authority), and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of acceptance used the base of power, legitimate (authority). The chi-square scores of the Legitimate (authority) group, ?² (5, N = 185) = 116.20 present no association with the scores of the 153 acceptance group, ?² (12, N = 185) = 91.65. Further, a crosstab analysis to determine association yields no statistical significance, ?² (60, N = 185) = 52.74, p = 0.73. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, Legitimate (Authority) and Acceptance; therefore, H0 4 was not rejected. Null Hypothesis Five There is no statistically significant association between the power base, reward, and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of acceptance used the base of power, reward. The chi-square scores of the reward group, ?² (4, N = 185) = 108.59 present no association with the scores of the acceptance group, ?² (12, N = 185) = 91.65. Further, a crosstab analysis to determine association yields no statistical significance, ?² (48, N = 185) = 38.13, p = 0.84. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, reward and acceptance; therefore, H0 5 was not rejected. Null Hypothesis Six There is no statistically significant association between the power base, coercive (discipline), and the attitude toward suicide, acceptance, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred 154 attitude toward suicide of acceptance used the base of power, coercive (discipline). The chi-square scores of the coercive (discipline) group, ?² (5, N = 185) = 87.14 present no association with the scores of the acceptance group, ?² (12, N = 185) = 91.65. Further, a crosstab analysis to determine association yields no statistical significance, ?² (60, N = 185) = 66.01, p = 0.28. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, reward and acceptance; therefore, H0 6 was not rejected. Null Hypothesis Seven There is no statistically significant association between the power base, information, and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of condemnation used the base of power, information. The chi-square scores of the information group, ?² (3, N = 185) = 99.43 present no association with the scores of the condemnation group, ?² (12, N = 185) = 102.61. Further, a crosstab analysis to determine association yields no statistical significance, ?² (36, N = 185) = 34.93, p = 0.52. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, information and condemnation; therefore, H0 7 was not rejected. Null Hypothesis Eight There is no statistically significant association between the power base, 155 expertise, and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of condemnation used the base of power, expertise. The chi-square scores of the expertise group, ?² (5, N = 185) = 119.70 present a statistically significant association with the scores of the condemnation group, ?² (12, N = 185) = 102.61. Further, a crosstab analysis to determine association yields statistical significance, ?² (60, N = 185) = 119.37, p = 0.000. Results of the Pearson Chi-Square analysis indicate statistically significant (p < 0.05) associations between variables, expertise and condemnation; therefore, H0 8 was rejected. Null Hypothesis Nine There is no statistically significant association between the power base, referent (goodwill), and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of condemnation used the base of power, referent (goodwill). The chi-square scores of the referent (goodwill) group, ?² (5, N = 185) = 112.76 present no association with the scores of the condemnation group, ?² (12, N = 185) = 102.61. Further, a crosstab analysis to determine association yields no statistical significance, ?² (60, N = 185) = 62.54, p = 0.38. Results of the Pearson Chi-Square analysis indicate no statistically 156 significant associations between variables, referent (goodwill) and condemnation; therefore, H0 9 was not rejected. Null Hypothesis Ten There is no statistically significant association between the power base, legitimate (authority), and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of condemnation used the base of power, legitimate (authority). The chi-square scores of the legitimate (authority) group, ?² (5, N = 185) = 116.20 present no association with the scores of the condemnation group, ?² (12, N = 185) = 102.61. Further, a crosstab analysis to determine association yields no statistical significance, ?² (60, N = 185) = 60.87, p = 0.44. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, legitimate (authority) and condemnation; therefore, H0 10 was not rejected. Null Hypothesis Eleven There is no statistically significant association between the power base, reward, and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of condemnation used the base of power, reward. The chi-square scores of the reward group, ?² (4, N = 185) = 108.59 present no association with the scores of 157 the condemnation group, ?² (12, N = 185) = 102.61. Further, a crosstab analysis to determine association yields no statistical significance, ?² (48, N = 185) = 44.51, p = 0.61. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, reward and condemnation; therefore, H0 11 was not rejected. Null Hypothesis Twelve There is no statistically significant association between the power base, coercive (discipline), and the attitude toward suicide, condemnation, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of condemnation used the base of power, coercive (discipline). The chi-square scores of the coercive (discipline) group, ?² (5, N = 185) = 87.14 present no association with the scores of the condemnation group, ?² (12, N = 185) = 102.61. Further, a crosstab analysis to determine association yields no statistical significance, ?² (60, N = 185) = 54.59, p = 0.67. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, reward and condemnation; therefore, H0 12 was not rejected. Null Hypothesis Thirteen There is no statistically significant association between the power base, information, and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to 158 assess the degree to which Deans of Students with the preferred attitude toward suicide of prevention used the base of power, information. The chi-square scores of the information group, ?² (3, N = 185) = 99.43 present no association with the scores of the prevention group, ?² (10, N = 185) = 89.16. Further, a crosstab analysis to determine association yields no statistical significance, ?² (30, N = 185) = 26.08, p = 0.67. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, information and prevention; therefore, H0 13 was not rejected. Null Hypothesis Fourteen There is no statistically significant association between the power base, expertise, and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of prevention used the base of power, expertise. The chi-square scores of the expertise group, ?² (5, N = 185) = 119.70 present a statistically significant association with the scores of the prevention group, ?² (10, N = 185) = 89.16. Further, a crosstab analysis to determine association yields statistical significance, ?² (50, N = 185) = 74.36, p = 0.01. Results of the Pearson Chi- Square analysis indicate statistically significant (p < 0.05) associations between variables, expertise and prevention; H0 14 was rejected. Null Hypothesis Fifteen There is no statistically significant association between the power base, 159 referent (goodwill), and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of prevention used the base of power, referent (goodwill). The chi-square scores of the referent (goodwill) group, ?² (5, N = 185) = 112.76 present no association with the scores of the prevention group, ?² (10, N = 185) = 89.16. Further, a crosstab analysis to determine association yields no statistical significance, ?² (50, N = 185) = 47.89, p = 0.56. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, referent (goodwill) and prevention; therefore, H0 15 was not rejected. Null Hypothesis Sixteen There is no statistically significant association between the power base, legitimate (authority), and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of prevention used the base of power, legitimate (authority). The chi-square scores of the legitimate (authority) group, ?² (5, N = 185) = 116.20 present a statistically significant association with the scores of the prevention group, ?² (10, N = 185) = 89.16. Further, a crosstab analysis to determine association yields statistical significance, ?² (50, N = 185) = 89.01, p = 0.001. Results of the Pearson Chi-Square analysis 160 indicate statistically significant (p < 0.05) associations between variables, legitimate (authority) and prevention; therefore, H0 16 was rejected. Null Hypothesis Seventeen There is no statistically significant association between the power base, reward, and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of prevention used the base of power, reward. The chi-square scores of the reward group, ?² (4, N = 185) = 108.59 present a statistically significant association with the scores of the prevention group, ?² (10, N = 185) = 89.16. Further, a crosstab analysis to determine association yields statistical significance, ?² (40, N = 185) = 57.20, p = 0.04. Results of the Pearson Chi- Square analysis indicate statistically significant (p < 0.05) associations between variables, reward and prevention; therefore, H0 17 was rejected. Null Hypothesis Eighteen There is no statistically significant association between the power base, coercive (Discipline), and the attitude toward suicide, prevention, in Deans of Students who are members of NASPA. A one-sample chi-square test was conducted to assess the degree to which Deans of Students with the preferred attitude toward suicide of prevention used the base of power, coercive (discipline). The chi-square scores of the coercive (discipline) group, ?² (5, N = 185) = 87.14 present no association with the scores of the prevention 161 group, ?² (10, N = 185) = 89.16. Further, a crosstab analysis to determine association yields no statistical significance, ?² (50, N = 185) = 44.18, p = 0.70. Results of the Pearson Chi-Square analysis indicate no statistically significant associations between variables, coercive (discipline) and prevention; therefore, H0 18 was not rejected. An independent samples t-test was applied to null hypotheses 19 and 20 to determine if significant differences were present between the independent variable, gender and the dependent variables, bases of power and attitudes toward suicide, respectively. Null Hypothesis Nineteen There is no statistically significant difference between the gender and the six bases of power in Deans of Students who are members of NASPA. This hypothesis aims to determine if there is a significant difference between the means of the independent variable of gender, specifically males (n = 89) and females (n = 95), and the bases of power or dependent variables of information, expertise, referent (goodwill), legitimate (authority), reward, and coercive (discipline). There is no statistically significant difference between gender and power base, information. The scores of the male group (M = 1.43, SD = 0.72) are not significantly different from the female group (M = 1.32, SD = 0.77): t (184) = 1.00; p = 0.31 (two-tailed) for the base of power, information. Equal variances 162 assumed, results of the independent samples t-test indicate no significant differences between the means; therefore, H0 19 was not rejected. There is no statistically significant difference between gender and power base, expertise. The scores of the male group (M = 2.58, SD = 1.07) are not significantly different from the female group (M = 2.51, SD = 0.99): t (184) = 0.52; p = 0.60 (two-tailed) for the base of power, expertise. Equal variances assumed, results of the independent samples t-test indicate no significant differences between the means; therefore, H0 19 was not rejected. There is no statistically significant difference between gender and power base, referent (goodwill). The scores of the male group (M = 3.19, SD = 1.11) are not significantly different from the female group (M = 3.25, SD = 1.05): t (184) = -0.38; p = 0.70 (two-tailed) for the base of power, referent (goodwill). Equal variances assumed, results of the independent samples t-test indicate no significant differences between the means; therefore, H0 19 was not rejected. There is no statistically significant difference between gender and power base, legitimate (authority). The scores of the male group (M = 2.73, SD = 0.97) are not significantly different from the female group (M = 2.66, SD = 1.06): t (184) = 0.44; p = 0.65 (two-tailed) for the base of power, legitimate (authority). Equal variances assumed, results of the independent samples t-test indicate no significant differences between the means; therefore, H0 19 was not rejected. There is no statistically significant difference between gender and power base, reward. The scores of the male group (M = 2.21, SD = 0.92) are not 163 significantly different from the female group (M = 2.35, SD = 0.85): t (184) = -1.02; p = 0.30 (two-tailed) for the base of power, reward. Equal variances assumed, results of the independent samples t-test indicate no significant differences between the means; therefore, H0 19 was not rejected. There is no statistically significant difference between gender and power base, coercive (discipline). The scores of the male group (M = 1.63, SD = 1.13) are not significantly different from the female group (M = 1.62, SD = 1.13): t (184) = 0.05; p = 0.96 (two-tailed) for the base of power, coercive (discipline). Equal variances assumed, results of the independent samples t-test indicate no significant differences between the means; therefore, H0 19 was not rejected. Null Hypothesis Twenty There is no statistically significant difference between gender and attitude toward suicide in Deans of Students who are members of NASPA. This hypothesis aims to determine if there is a significant difference between the means of the independent variable of gender, specifically males (N = 89) and females (N = 95), and the attitudes toward suicide or dependent variables of acceptance, condemnation, and prevention. There is no statistically significant difference between gender and attitude toward suicide, acceptance. The scores of the male group (M = 9.11, SD = 2.66) are not significantly different from the female group (M = 9.17, SD = 2.58): t (184) = -0.17; p = 0.86 (two-tailed) for the attitude toward suicide, acceptance. 164 Equal variances assumed, results of the independent samples t-test found no significant differences between the means; therefore, H0 20 was not rejected. There is no statistically significant difference between gender and attitude toward suicide, condemnation. The scores of the male group (M = 8.13, SD = 2.46) are not significantly different from the female group (M = 8.37, SD = 2.33): t (184) = -0.69; p = 0.49 (two-tailed) for the attitude toward suicide, condemnation. Equal variances assumed, results of the independent samples t- test found no significant differences between the means; therefore, H0 20 was not rejected. There is no statistically significant difference between gender and attitude toward suicide, prevention. The scores of the male group (M = 6.75, SD = 2.23) are not significantly different from the female group (M = 6.96, SD = 2.09): t (184) = -0.67; p = 0.50 (two-tailed) for the attitude toward suicide, prevention. Equal variances assumed, results of the independent samples t-test found no significant differences between the means; therefore, H0 20 was not rejected. Summary of Significant Findings The analysis of the first 18 hypotheses produced four associations of statistical significance. In three of the associations, Deans of Students with the preferred attitude toward suicide of prevention report a preference toward the power bases, expertise, legitimate (authority), and reward. The fourth association indicates that Deans of Students with the preferred attitude toward suicide of condemnation report a preference toward the power base, expertise. Results of 165 associations between variables, attitudes toward suicide and bases of power are presented in Table 2. An overview of the final analysis of hypotheses 19 and 20 yield no associations of statistical significance. Hypotheses 19 and 20 are not rejected and no statistical differences exist between gender and attitudes toward suicide and gender and bases of power. Challenges with Instrumentation Objections related to the format, administration, and scoring of the PBI are worthy of discussion. The ipsative format of the PBI concerned eight participants. These respondents said the forced-choice format of the PBI was too confining, it did not reflect their individual beliefs about power, and they did not agree with either option presented in the survey. As a result, these eight respondents elected not to participate in this study. The administration and scoring of the PBI presented unique challenges. In its present format of 30 paired statements, the PBI is long and repetitive. The 30 paired statements were reduced by the researcher to fifteen for brevity and simplicity. Reducing the number of statements present in the inventory also reduced the number of times the power base was used in each paired statement from ten to five. As a result, scoring the inventory was modified by CCP, Inc.’s research associates, and CCP, Inc.’s research associates reported that these modifications might impact the reliability and validity of the scores. Earlier in this chapter, details of the group scores of the PBI were reported under section titled, Participants’ Responses. Computation of the group 166 Table 2 Results of Chi-Square Tests for Hypotheses 1 – 18 Variable Value df p Acceptance Information 33.51 36 0.58 Expertise 66.38 60 0.25 Referent (Goodwill) 50.86 60 0.79 Legitimate (Authority) 52.74 60 0.73 Reward 38.13 48 0.84 Coercive (Discipline) 66.01 60 0.27 Condemnation Information 34.93 36 0.51 Expertise 19.37 60 0.000* Referent (Goodwill) 62.54 60 0.38 Legitimate (Authority) 60.87 60 0.68 Reward 44.51 48 0.61 Coercive (Discipline) 54.59 60 0.67 Prevention Information 26.08 30 0.67 Expertise 74.36 50 0.01* Referent (Goodwill) 47.89 50 0.55 Legitimate (Authority) 89.01 50 0.001* 167 Table 2 Results of Chi-Square Tests for Hypotheses 1 – 18 (continued) Variable Value df p Reward 57.20 40 0.04* Coercive (Discipline) 44.18 50 0.70 Note. * p < 0.05. 168 scores were obtained by following the instructions for scoring the PBI provided by the inventory authors Thomas (1985) and Thomas. However, the procedures for computing scores on the PBI are intended to be done for individual scores, not group scores. Therefore, an understanding of how the PBI was intended to be scored must be considered when interpreting the group findings mentioned earlier in this chapter. Conclusion The purpose of this research was to examine the associations between the power base theory (French & Raven, 1968) and the three factor model of attitudes toward suicide (Salander-Renberg et al., 2008) in Deans of Students. In addition, analysis to determine associations between gender and attitudes toward suicide and gender and bases of power were assessed. Findings indicate that Deans of Students who believe suicide is preventable perceive their base of power as expertise, legitimate (authority), or reward. Deans of Students with the opinion of condemnation toward suicide tend to use the power base, expertise. No statistically significant associations between gender and attitudes toward suicide and gender and bases of power exist. CHAPTER 5: DISCUSSION, IMPLICATIONS AND RECOMMENDATIONS Introduction The purpose of this study was to investigate the research question: What is the association between attitudes toward suicide and bases of power in Deans of Students (DOS)? The conceptual framework that addresses the research question consists of French and Raven’s (1968) social power base theory (the independent variable) and Salender-Renberg and Jacobbson’s (2003) three factor model of attitudes towards suicide (the dependent variable). This chapter includes: (a) a summary of the participants’ demographics and major findings of this study, (b) a discussion of the findings and the conceptual framework, and (c) implications for educational leaders, recommendations for future research, and limitations of the study. Participant Demographics The participants in this study are members of the National Association of Student Personnel Administrators (NASPA) who perform the duties of DOS. Approximately half of the individuals eligible to participate in this study elected to respond to the survey instrument. Responses from men and women are almost evenly split with slightly more women participating than men. Half of the participants said they earned a doctorate while almost the entire other half earned a master’s degree. Approximately an equal representation of gender groups and highest degree earned categories is present. 170 Major Findings of the Study The findings of the study reveal DOS perceive their position of authority, expertise in student affairs, and an intrinsic incentive to save a life, is associated with their opinion that suicide can be prevented. In addition, DOS perceive their expertise in student affairs to be associated with their opinion that suicide should be condemned. DOS also believe people have the right to take their own life, and DOS say they understand why an individual might want to commit suicide. Finally, gender appears to have no impact on DOS’ attitudes toward suicide or level of authority. Discussion of the Results Means for identifying students who might commit suicide are the responsibility of DOS as leaders in student affairs. Therefore, an understanding of DOS attitudes toward suicide and bases of power may be useful in assessing current DOS’ motivation and ability to effectively identify students who might commit suicide. The results of this study indicate four significant findings. First, DOS think they possess a level of professional expertise which appears to be associated with their belief that suicide can be prevented. Second, DOS think their position of authority which appears to be associated with their belief that suicide can be prevented. Third, DOS think the presence of reward power appears to be associated with their belief that suicide can be prevented. Fourth, DOS think they possess a level of professional expertise which appears to be associated with their belief that suicide should be condemned. This discussion of 171 the results briefly discusses each finding, connects the finding to the literature, and interprets the finding. Finding 1: Expertise Power and Prevention of Suicide Leaders who perceive to influence others through the use of expertise power think their subordinates, like student affairs personnel, comply with the leaders’ requests because the leader is viewed as the most proficient, skilled individual in the unit (Elias, 2008). In essence, the boss knows what is best for the organization because the boss is recognized as the most knowledgeable person in the unit. Finding 1 reveals that DOS believe they possess a level of professional expertise which appears to be associated with their belief that suicide can be prevented. It is plausible DOS view themselves as experts in student affairs because they usually possess a high level of education and a substantial number of years of experience in their field. For example, the participants in this study report being highly educated, 50% earned a doctorate and 45% earned a master’s. In addition, DOS are often promoted to this position after obtaining a wealth of experience in student affairs. Komives et al. (2003) suggest that DOS’ think their knowledge and expertise in matters related to student’s well-being is exceptional; therefore, DOS would think suicide could be preventable. It would be reasonable to assume DOS believe the responsibility of preventing suicide rests with them. In addition, as highly educated professionals, DOS’ knowledge about the complex social and emotional realities of student life on campus is thought to be 172 extensive (Sandeen & Barr, 2006). DOS are fully aware of the societal stressors associated with being a college student. Therefore, it is no surprise to find that DOS report understanding why an individual, like an overly stressed student, might commit suicide. This familiarity with the realities of student life and an understanding of typical adolescent behaviors displayed by university students, in many ways, contributes to DOS affirmation that suicide might be a real option for some students. It is reasonable to suggest that DOS possess a realistic understanding of the psychological/emotional stresses of students face. Therefore, expertise in students’ psychological/emotional development might have some association with DOS’ perceptions that suicide can be prevented. Finding 2: Legitimate (Authority) Power and Prevention of Suicide Leaders who believe they have legitimate power feel subordinates, like student affairs personnel, comply with their requests by virtue of their position of authority (Elias, 2008). In effect, the boss possesses a level of power because he or she is the boss. Finding 2 reveals that DOS believe they possess legitimate power which appears to be associated with their belief that suicide can be prevented. The title Dean of Students is considered a leadership position within the division of student affairs (Nuss, 2003). DOS portrayal of legitimate power exists by virtue of the hierarchical structure of the university system. Therefore, this appointed position of authority could influence their perceived duty to prevent a 173 student from ending his or her life. As a result, it would be fair to say that DOS think the responsibility of preventing suicide rests with them. In addition, DOS position of authority might contribute to their keen awareness of students’ rights as individuals. DOS tend to value student autonomy which they feel should be respected and nurtured in the university setting (Nuss, 2003). Fostering the development of student autonomy is deemed an important element of the university’s culture especially since the demise of in loco parentis (Komives et al, 2003). The right to commit suicide might be perceived as the prerogative of the student, according to DOS. Therefore, while it is reasonable to assume that DOS believe individuals have the right to die by suicide, DOS perceived authority appears to be associated with their belief that suicide can be prevented, the right to commit suicide notwithstanding. Finding 3: Reward Power and Prevention of Suicide Leaders who utilize reward power believe subordinates, like student affairs personnel, comply with their requests because the subordinate is receiving compensation for a job well done (Elias, 2008). At heart, the boss feels subordinates follow through with his or her requests because the subordinates are being compensated for demonstrating compliance. Finding 3 reveals that DOS believe they possess reward power which appears to be associated with their belief that suicide can be prevented. At first glance, this association is unclear. The act of rewarding subordinates for complying with the boss does not seem to relate to the belief that suicide can be preventable. 174 However, one might assert that an intrinsic reward system in the university workplace exists. For the purposes of this discussion, an intrinsic reward is a fundamental or natural incentive to want to prevent suicide, an example of a positive reward. Results of this study indicate that DOS believe humans have a duty to try to prevent suicide. Sandeen and Barr (2006) confirmed that DOS perceive a duty to facilitate the maintenance of the safety and well-being of students which might include the supervision of suicide prevention efforts. The compensation or reward for this perceived duty is the sentiment that caring about preventing suicide could be interpreted by others as noble and admirable. Righteous qualities, including the belief that it is one’s duty to prevent suicide, could be interpreted as an intrinsic reward. DOS appear to recognize the intrinsic reward is the perceived duty to prevent suicide because it is viewed by the university community as the dignified thing to do. Finding 4: Expertise Power and Condemnation of Suicide Condemnation of suicide is the opinion that suicide is a forbidden act (Salander-Renberg et al., 2008). Finding 4 reveals that DOS believe they possess a level of expertise power which appears to be associated with their belief that suicide should be condemned. It does not seem probable that an association exists between DOS expert knowledge and belief that suicide should be condemned. DOS are cognizant of the challenges and stresses of student life (Komives et al., 2003). One would think this awareness of the pressures of student life and knowledge about suicide would result in an opinion that favors an 175 understanding of suicide. Individuals who identify with the realities of one’s decision to commit suicide generally do not condemn suicide. A closer examination of the condemnation statements on the Attitudes toward Suicide (ATTS) questionnaire offers an explanation for this association. A significant number of DOS disagree or are undecided about condemnation statements, Suicide can never be justified and I do not understand how people take their lives. Disagreeing or feeling undecided about these statements might mean respondents accept suicide as a real possibility for some individuals. In addition, the context of the suicidal behavior might impact an individual’s attitudes toward suicide. For example, suicide bombers who typically kill themselves for a cause engage in this behavior for very different reasons than the university student who perceives life should end because of perceived poor performance in the classroom. Hence, a measurement error might exist in this finding as it equivocates unlike situations. In disagreeing and feeling undecided about these condemnation statements, DOS are reporting that suicide could be an acceptable option for individuals who feel suicide is the only solution to their problems. In addition, DOS probably accept the unfortunate certainty that some individuals will commit suicide despite DOS efforts to prevent it. Therefore, it does not seem accurate to report that DOS’ expertise power appears to be associated with their belief that suicide should be condemned. It would be more precise to say that DOS believe 176 suicide could be justified, and on the whole, they appear to understand why a person might commit suicide. The literature concedes that individuals’ attitudes toward suicide, in general, are difficult to predict, remain complex, and lack consistency (Bayet, 1922; Gibbs, 1968; Ginn et al., 1988; McLaughlin, 2007; Platt, 1989; Sale et al., 1975). Morgan et al. (1996) found that professionals with knowledge about suicide prevention thought they were more adept at preventing suicide. Ultimately, the responsibilities associated with this position of authority, the degree of expertise these leaders bring to this position, and a certain amount of intrinsic incentive, either positive or negative, appears to influence DOS’ confidence that suicide can be prevented. These findings confirm what others have determined. Attitudes toward suicide are not exact, but instead, are complex, ever-changing, and have the potential to be influenced by some powers. Conceptual Framework Inquiries about DOS attitudes toward suicide and their perceived authority on campus are appropriate to consider since DOS are generally responsible for the safety of students while they are on campus. The conceptual framework of the three factor model of attitudes toward suicide and the six bases of social power is used to investigate these matters. The following discussion provides interpretations of the conceptual framework for this study. 177 The death of a student by suicide on a university campus is difficult to predict but almost always produces intense reactions from the university community. For example, after five students at Cornell University committed suicide by the end of the 2010 academic year, conversations alluding to Cornell University’s reputation as a ‘suicide school’ erupted (Epstein, 2010). There is an extensive amount of literature that discusses the psychological emotions, cultural or societal beliefs, and religious perspectives which impact individuals’ attitudes toward suicide. However, attempts to measure opinions about suicide continue to be a challenge for researchers. And like others, Salander-Renberg and Jacobsson (2003) and Salander-Renberg et al. (2008) concede that measuring opinions about suicide can be difficult because attitudes toward suicide are somewhat unpredictable and inconsistent. The social power theory (French & Raven, 1968) has been evaluated at length (Ambur, 2000; Elias, 2008; Hayden, 2000; Hinkin & Schriesheim, 1994; Pierro et al., 2008). Researchers define power as the means by which an individual convinces another person to comply with his or her wishes. In many instances, efficient leaders use their power and influence to win over their subordinates. Raven (1993) asserts that power is complex, ever changing, and situational. Power appears to be multifaceted in a way that includes every member of an organizational unit. Each member of an organization has the opportunity to seize some level of authority. 178 The use of this conceptual framework yielded significant findings supported in the literature. DOS possess a level of authority based on their position at the institution and the amount of experience they bring to the organization (Appleton et al., 1978; Komives et al., 2003; Sandeen & Barr, 2006; Schwartz, 2002). One might suggest that DOS’ power might be associated with their belief that suicide can be prevented. However, findings discussed in this review of the literature indicate that the study of suicide and the study of power are complex and sometimes contradictory. Results from this study confirm what is being discussed in the literature. The association between attitudes toward suicide and some bases of power appears to exist. Evidence of the association expertise power and legitimate power has on DOS desire to prevent suicide is apparent; however, evaluation of this framework requires further investigation. Implications for Educational Leaders Outcomes of this study have several implications for educational leaders. Overall, DOS’ position of power and professional expertise appears to be associated with their belief that suicide can be prevented. Accordingly, as leaders in student affairs, DOS may wish to consider managing the institution’s discussions, policies, and practices related to identifying students at risk for committing suicide. Based on the outcomes of this research, implications for educational leaders are discussed. DOS should consider conducting university wide discussions about suicide. University wide conversations about suicide will allow individuals’ to 179 explore their opinions about suicide so DOS can better gauge the community’s knowledge about suicide. Researchers at The Jed Foundation (2006) assert that university leaders do not spend enough time engaging students, faculty, and staff in discussions about suicide. Findings in this study suggest that DOS believe suicide can be prevented; therefore, dialogues initiated by DOS have the potential to keep deans abreast of individuals’ perceptions of suicide for the purposes of improving suicide prevention efforts. It seems appropriate for DOS to be active, contributing members, of the university crisis intervention team. The establishment of university crisis intervention teams is the result of the increase in violence to self and others on university campuses. The purpose of this team is to respond to campus emergencies including ones involving students who might attempt suicide. Findings in this study suggest DOS believe suicide can be prevented. As one of the primary administrators responsible for the safety and well-being of students, perhaps DOS should consider being one of the lead members of this team. The initial point-of-contact for students in crisis could possibly be DOS. In addition, DOS may wish to think about organizing a committee of university staff support service professionals who are trained to assist the dean with aiding students during a crisis situation. Although, the results of this study indicate DOS possess the authority and expertise to prevent suicide, the diversity and complexity of student crisis situations might require DOS to utilize the skills and abilities of staff with experience responding to specific student crisis issues. 180 As a member of the university crisis intervention team with knowledge and expertise in matters related to recognizing students who display suicidal tendencies, DOS should consider leading the efforts to identify the best practices for preventing suicide. The literature indicates that university administrators struggle to identify students at risk for committing suicide (The Jed Foundation, 2006; Kitzrow, 2003; Silverman, 1993; Suicide Prevention Resource Center, 2004). The findings in this study propose that DOS hold the expertise, position of authority, and belief that suicide can be prevented. Researchers advocate for DOS to lead the suicide prevention efforts on their campus because they are considered the authority on campus with the knowledge to prevent suicide (Suicide Prevention Resource Center, 2004). Perhaps DOS could implement university wide social support programs designed to identify and respond to students prone to suicidal behaviors. DOS should consider offering professional development training sessions, specifically for faculty, to inform them of the procedures for identifying students at risk for committing suicide. Research conducted by The Jed Foundation (2006) reported that faculty was uncertain about the protocol for identifying students who might take their life. DOS position of authority and skill in matters related to student life suggests DOS might be the logical choice for facilitating suicide prevention training sessions. Advocacy for additional funding and resources for staff that operate university counseling centers appears justified and might be initiated by DOS. 181 These centers generally play a large part in assisting student in crisis, yet counseling center professionals report they lack sufficient staff and resources to respond to the increasing numbers of students in crisis (Cook, 2007). DOS position of leadership within the organizational structure of the university provides them the authority to access additional funding. Finally, it seems appropriate for DOS to investigate ways they can create a campus community that encourages students to seek help prior to the onset of suicidal behaviors. Cook (2007) and Westefeld et al. (2005) found that students attach a negative stigma to asking for help when they present suicidal behaviors. DOS position of authority and belief that suicide can be prevented suggests DOS appear to have the power to influence students’ perceptions of asking for help. To start, DOS could initiate a series of focus groups, comprised of members of the student body, faculty, and staff, for the purposes of discussing students’ thoughts about asking for help. Following an analysis of the data from the focus groups, perhaps DOS could take the lead in educating students, faculty, and staff about the benefits to seeking help. These recommendations are suggested starting points for creating a comprehensive program aimed at removing the negative stigma students attach to seeking help. Recommendations for Future Research The methodology of this study surveyed a large population using a short survey instrument in order to obtain a satisfactory response rate. Findings are broad-based but relevant to a specific population. However, there are limits to the 182 ability of the survey’s design to ascertain a circumspect understanding of the association between attitudes toward suicide and bases of power. The following suggestions for future research support the outcomes of this study. Conduct a qualitative analysis of opinions about suicide and perceptions of power with university administrators at a single institution as the topics of suicide and power are complex and elicit strong emotional feelings from individuals. A qualitative study would provide a more in-depth analysis of these issues and expand on the findings presented in this study. The need for future research here is confirmed by DOS who sent emails to the researcher stating that the survey instruments used in this study did not accurately capture their diverse perceptions of attitudes toward suicide and bases of power. An investigation of the influence race/ethnicity has on attitudes towards suicide and perceived authority in DOS is warranted. Colleges and universities are seeing an influx of professionals from diverse cultural backgrounds. It is unknown what impact race/ethnicity has on attitudes toward suicide and levels of authority in DOS from diverse racial or ethnic backgrounds. In addition, as this investigation did not consider matters of race/ethnicity, religiosity, and other demographic characteristics, future work could elucidate the influence of these on attitudes toward suicide and levels of authority in DOS. A study exploring the association between DOS age and perceptions about suicide prevention is recommended. Inconsistencies in the literature regarding the impact of an individuals’ age on their perception of suicide 183 prevention exist. For example, Salander-Renberg et al. (2008) found that older adults described themselves as pro-suicide prevention when compared with their younger counterparts; whereas, Mofidi et al. (2008) found younger adults to be more pro-suicide prevention than the older adults. Exploring the association between DOS age and perceptions about suicide prevention would further contribute to this existing disparity in the literature. Conduct a meta-analysis using existing data to extract information about the suicide prevention efforts employed by administrators in student affairs would be useful towards identifying some of the best practices in suicide prevention. Current practices regarding the identification of students at risk for committing suicide have been studied frequently, but usually as single-site studies. Limitations of the Study Participants who responded to the survey instrument are members of NASPA and not all DOS elect to join NASPA. Therefore, results of the study are only generalizable to DOS who are members of NASPA. Additionally, modifications to the instrumentation used in the study were made with the intention of maximizing the opportunity to achieve a good response rate. Scoring modifications made to the PBI meant the research associates at CPP, Inc. could not certify the reliability and validity of the PBI scores. Future work can confirm the reliability and validity of the shortened PBI. 184 Conclusion This study investigated the association between attitudes toward suicide and bases of power in DOS who are members of NASPA. The results of this study indicate that DOS perceive their expertise in student affairs and their position of authority appears to be associated with their belief that suicide is preventable. In addition, an intrinsic incentive related to a perceived duty to prevent suicide exists among DOS. Responsible for the development of the whole student, DOS play a fundamental role in the methods used to identify students at risk for committing suicide (Jed Foundation, 2006). At present, inconsistencies regarding the methods used by DOS to respond to students who consider committing suicide exist (Francis, 2003; The Jed Foundation, 2006; Kitzrow, 2003; Suicide Prevention Resource Center, 2004). Yet, findings indicate that power in the form of position of authority and level of expertise in matters related to student affairs appears to be associated with DOS’ belief that suicide can be prevented. Accordingly, as leaders in student affairs, DOS possess the knowledge and skill needed to lead the efforts required to manage the institution’s practices related to identifying students at risk for committing suicide. Just one suicide on a campus in a given year is a tragedy; yet, suicide remains the second leading cause of death among college students. DOS’ preferences for legitimate, expertise, and reward power demonstrates they are well poised to take charge of this area, given their authority on campus and 185 knowledge of student development. Furthermore, DOS condemnation of suicide can fuel conversations with members of the university community regarding opinions about the context of suicide, and circumstances when suicide might be justified. REFERENCES Adair, S. (2008). Status and solidarity: A reformulation of early Durkheimian theory. Sociological Inquiry, 78(1), 97 – 120. Ajzen, I. (1985). From intentions to actions: A theory of planned behavior, in Action Control: From Cognition to Behavior (eds. J. Kuhl & J. Beckmann), Springer, Heidelberg. Ambur, O. (2000). Reconsidering the higher-order legitimacy of French and Raven’s bases of social power in the information age. University of Maryland University College. Retrieved November 29, 2008, from http://ambur.net/French&Raven.htm. American College Health Association. (2001). National College Health Assessment: Aggregate report, Spring 2000. Baltimore: Author. American College Health Association. (2004). National college health assessment spring 2004 reference group data report. Journal of American College Health, 54(4), 201-211. American Psychiatric Association (APA). (2003). Practice guideline for the assessment and treatment of patients with suicidal behaviors. American Journal of Psychiatry, 160(supplement, 1-60). Anderson, A. L., Lester, D., & Rogers, J. R. (2008). A psychometric investigation of the suicide opinion questionnaire. Death Studies, 32, 924-936. Appelbaum, P. S. (2006). Depressed? Get Out?: Dealing with suicidal students on college campuses. Law & Psychiatry, 57, 914-916. 187 Appleton, J. R., Briggs, C. M., & Rhatigan, J. J. (1978). Pieces of Eight: The rites, roles, and styles of the Dean. Portland, Oregan: NASPA Institute of Research and Development. Associated Press, The. (2008, November 21). Retrieved December 7, 2008 from http://www.nbcaugusta.com/news/national/34914539.html. Bachman, J. G., Smith, C. G., & Slesinger, J. A. (1966). Control, performance, and satisfaction: An analysis of structural and individual effects. Journal of Personality and Social Psychology, 4, 127-136. Barnes, R. E. (1963). So you want to be a Dean of Students. Peabody Journal of Education, 41(1), 3 – 9. Barr, M. J., & Keating, L. A. (1985). Developing effective student services programs. San Francisco: Jossey-Bass. Barrios, L. C., Everett, S. A., Simon, T. R., & Brener, N. D. (2000). Suicide ideation among U.S. college students. College Health, 48, 229-233. Basi, R. K. (1991). WWW response rates to socio-demographic items. Journal of the Market Research Society, 41(4), 397-498. Bass, B. M. (1990). Bass and Stogdill’s handbook of leadership (3rd ed.). New York: Free Press. Bayet, A. (1922). Le Suicide et la Morale. Paris: Felix Alcan. Berk, B. B. (2006). Macro-micro relationships in Durkheim’s analysis of egoistic suicide. Sociological Theory, 24, 58 – 80. 188 Biggert, N. W., & Hamilton, G. G. (1984). The power of obedience. Administrative Science Quarterly, 29, 540-549. Blue, J. F., Jr. (1972). The assumptions and beliefs of selected academic community members with regard to the role of the dean of students in North Carolina. Unpublished doctoral dissertation, University of North Carolina, Chapel Hill. Bosnjak, M. & Tuten, T. L. (2001). Classifying response behaviors in web-based survey. Journal of Computer-Mediated Communication, 6(3). Retrieved February 22, 2010, from http:/www.ASCUSC.ORG/JCMC/VOL16/ISSUE3/BOZNJAK.HTML. Bost, J. M., & Ballou, R. A. (1992). The issue of suicide: Policy response for a small liberal arts university. TCA Journal, 20, 33-39. Brener, N. D., Hassan, S. S., & Barrios, L. C. (1999). Suicidal ideation among college students in the United States. Journal of Consulting and Clinical Psychology, 67, 1004-1008. Carson, P. P., Carson, K. D., Knight, E. L., & Roe, C. W. (1995). Power in organizations: A look through the TQM lens. Quality Progress, 28(11), 73- 79. Cartwright, D. (1959). Power: A neglected variable in social psychology. In D. Cartwright (Ed.), Studies in social power (pp. 1-14). Ann Arbor, MI: Institute for Social Research. 189 Cartwright, D. (1965). Influence, leadership, and control. In J. G. March (Ed.), Handbook of organizations. (pp. 1-47). Chicago: Rand McNally. Center for Disease Control and Prevention. (1995). National Center for Chronic and Health Promotion, Youth Risk Surveillance: National College Health Risk Survey. Center for Disease Control and Prevention. (2006). Leading causes of death reports: 1999-2003. National Center for Injury Prevention and Control. Retrieved April 12, 2008 from http://webappa.cdc.gov/sasweb/ncipc/leadcaus10.html. Clark, T. A. (1921). Discipline and the derelict: Being a series of essays on some of those who tread the green carpet. New York: Macmillan. Clemons, J. T. (1990). Perspectives on Suicide. Louisville, Kentucky: Westminster/John Knox Press. Coch, L., & French Jr., J. R. P. (1968). Overcoming resistance to change. In D. Cartwright & A. Zander (Eds.), Group dynamics: research and theory (pp. 336-350). New York: Harper & Row, Publishers. Conner, K. R., Duberstein, P. R., Conwell, Y., & Seidlitz, L. (2001). Psychological vulnerability to completed suicide: A review of empirical studies. Suicide & Life-Threatening Behavior, 31(4), 367 – 386. Cook, L. J. (2007). Striving to help college students with mental health issues. Journal of Psychosocial Nursing, 45, 40-44. 190 Diekstra, R. F. W., & Kerkhof, A. J. F. M. (1998). Attitudes toward suicide: Development of a suicide attitude questionnaire (SUIATT). Current Issues of Suicidology, 462-476. Domino, G. (2005). Cross-cultural attitudes towards suicide: The SOQ and a personal odyssey. Archives of Suicide Research, 9, 107-122. Domino, G., Gibson, L., & Poling, S. (1980). Students’ attitudes towards suicide. Social Psychiatry, 15, 127-130. Domino, G., Moore, D., Westlake, L., & Gibson, L. (1982). Attitudes toward suicide. Journal of Clinical Psychology, 38, 257-262. Durkheim, E. (1951). Suicide: A study in sociology. New York: The Free Press. Early, K. E. (1992). Religion and suicide in the African-American community. Westport, CT: Greenwood Press. Eddy, J. P., Chen, C. H., & Ball, J. L. (1988). Historical development of the dean of students. The College Student Affairs Journal, 8(3), 6-9. Eddy, J. P., & Sharma, T. N. (1986). Student services: Development and referrals. In J. P. Eddy & R. S. Inchassi (Eds.), Higher Education Student Affairs Leadership: Programming, Budgeting, and Evaluating (pp. 42-45). Minneapolis, MN: Alpha. Elias, S. M. (2008). Fifty years of influence in the workplace: The evolution of the French and Raven power taxonomy. Journal of Management History, 14(3), 267-283. 191 Elliott, J. L. (2004). [Review of the book Power Base Inventory]. The Thirteenth Mental Measurement Yearbook. Lincoln, NE: Buros Institute of Mental Measurements. Retrieved April 15, 2009, from http://jproxy.lib.ecu.edu/login?url=http://search.ebscohost.com/login.aspx? direct=true&db=loh&AN=13181677&site=ehost-live. Epstein, J. (2010). A ‘suicide school’? Inside Higher Ed.com daily update. Retrieved March 16, 2010, from www.insidehighered.com/newsletter/html Erchul, W. P., Raven, B. H., & Whichard, S. M. (2001). School psychologist teacher perceptions of social power in consultation. Journal of School Psychology, 39(6), 483-497. Etzersdorfer, E., Vijayakumar, L., Schony, W., Grausgruber, A., & Sonneck, G. (1998). Attitudes towards suicide among medical students: Comparison between Madras (India) and Vienna (Austria). Social Psychiatry and Psychiatric Epidemiology, 33, 104-110. Festinger, L. (1954). Theory of social comparison processes. Human Relations, 7, 117-140. Francis, P. C. (2003). Developing ethical institutional policies and procedures for working with suicidal students on a college campus. Journal of College Counseling, 6, 114-123. French, J. R. P. (1956). A formal theory of social power. Psychological Review, 63(3), 181-194. 192 French J. R. P., & Raven, B. H. (1959). The bases of social power. In D. Cartwright (Ed.), Studies in social power (pp. 150-167). Ann Arbor, MI: Institute for Social Research. French Jr., R. P., & Raven, B. (1968). The bases of social power. In D. Cartwright & A. Zander (Eds.), Group dynamics: research and theory (pp. 259-269). New York: Harper & Row, Publishers. Freud, S. (1915). Instincts and their vicissitudes, in Standard Edition, 14, 109- 140. London: Hogarth Press. Freud, S. (1917). Mourning and melancholia, in Standard Edition, 14, 237-260. London: Hogarth Press. Freud, S. (1961). Mourning and melancholia. In J. Strachey (Ed. And Trans.), The standard edition of complete psychological works of Sigmund Freud (Vol. 14, pp.243-258). London: Hograth Press. (Original work published in 1917). Furr, S., Westefeld, J. S., Gaye, N., McConnell, G. N., & Marshall, J. (2001). Suicide and depression among college students: A decade later. Professional Psychology: Research and Practice, 32(1), 97-100. Fuse, T. (1997). Suicide, individual and society. Toronto, Canada: Canadian Scholars’ Press. Gane, M. (2005). Durkheim’s scenography of suicide. Economy and Society, 34(2), 223 – 240. 193 Gerda, J. J. (2006). Gathering together: A view of the earliest student affairs professional organizations. NASPA Journal, 43(4), 147 – 163. Gibbs, J. P. (Ed.). (1968). Suicide. New York: Harper-Row. Ginn, P. D., Range, L. M., & Hailey, B. J. (1988). Community attitudes towards childhood suicide and attempted suicide. Journal of Community Psychology, 16(2), 144-151. Graeff, P., & Mehlkop, G. (2007). When anomie becomes a reason for suicide: A new macro-sociological approach in the Durkheimian tradition. European Sociological Review, 23(4), 521 – 535. Grayson, P. (1994). Campus suicide: A memoir. Journal of College Student Psychotherapy, 9, 57-68. Gutierrez, P. M., Osman, A., Kopper, B. A., Barrios, F. X., & Bagge, C. L. (2000). Suicide risk assessment in a college student population. Journal of Counseling Psychology, 47, 403-413. Haas, A. P., Hendin, H., & Mann, J. J. (2003). Suicide in college students. The American Behavioral Scientist, 46(6), 1224-1240. Hasselback, P., Lee, K. I., Yang, M., Nichol, R., & Wigle, D. (1991). The relationship of suicide rates to sociodemographic factors in Canadian census divisions. Canadian Journal of Psychiatry, 36, 655-659. Hayden, W. D. (2000). The Power Theory of Leadership. Kent State University, Stark Campus. Retrieved November 29, 2008, from http://ambur.net/French&Raven.htm. 194 Hinkin, T., & Schriesheim, C. A. (1994). An examination of subordinate-perceived relationships between leader reward and punishment behavior and leader bases of power. Human Relations, 47(7), 779-811. Holmes, R. M. & Holmes, S. T. (2005). Suicide: Theory, practice, investigation. Thousand Oaks, California: Sage Publications. Hurwitz, J. I., Zander, A. F., & Hymovitch, B. (1953). Some effects of power on relations among group members. In D. Cartwright & A. F. Zander (Eds.), Group dynamics: Research and theory (pp. 483-492). Evanston, IL: Row, Peterson. Ivancevich, J. M. (1970). An analysis of control, bases of control, and satisfaction in an organizational setting. Academy of Management Journal, 13, 427- 436. Jed Foundation, The. (2006). Framework for developing institutional protocols for the acutely distressed or suicidal college student. New York, NY: The Jed Foundation. Kinnear, P. R., & Gray, C. D. (2010). PASW 17 Statistics Made Simple: Replaces SPSS Statistics 17. New York: Psychology Press. Kisch, J., Leino, E. V., & Silverman, M. M. (2005). Aspects of suicidal behavior, depression, and treatment in college students: Results from the spring 2000 national college health assessment survey. Suicide & Life- Threatening Behavior, 35, 3-13. 195 Kitzrow, M. (2003). The mental health needs of today’s college students: Challenges and recommendations. NASPA Journal, 41, 165-183. Komives, S., Woodard, Jr., D., & Associates. (2003). Student Services: A handbook for the profession (4th ed.). San Francisco: Jossey-Bass. Konick, L. C., & Gutierrez, P. M. (2005). Testing a model of suicide ideation in college students. Suicide & Life-Threatening Behavior, 35, 181-192. Kotter, J. P. (1985). Power and Influence: Beyond Formal Authority. New York: Free Press. Langhinrichsen-Rohling, J., Arata, C., Bowers, D., O’Brien, N., & Morgan, A. (2004). Suicidal behavior, negative affect, gender, and self-reported delinquency in college students. Suicide & Life-Threatening Behavior, 34, 255-266. Leary, T. (1957). Interpersonal diagnosis of personality. New York: Roland. Lees, J. & Stimpson, Q. (2008). A psychodynamic approach to suicide. In S. Palmer (Ed.), Suicide: Strategies and intervention for reduction and prevention (pp. 173-183). New York: Routledge. Leenaars, A. A. (1988). Suicide notes. New York: Human Sciences Press. Leenaars, A. A. (1989a). Suicide across the adult lifespan: An archival study. Crisis, 10, 132-151. Leenaars, A. A. (1989b). Are young adults’ suicides psychologically different from those of other adults? (The Shneidman Lecture). Suicide & Life Threatening Behavior, 19, 249-263. 196 Leenaars, A. A. (1993). Unconscious processes. In A. Leenaars (Ed.), Suicidology: Essays in honor of Edwin Shneidman (pp. 126-147). Northvale, NJ: Aronson. Leenaars, A. A. (1996). Suicide: A multidimensional malaise. (The Presidential Address). Suicide & Life-Threatening Behavior, 26, 221-236. Leenaars, A. A. (1999). Suicide across the adult lifespan: Replications and failures. Archives of Suicide Research, 5, 261-274. Leenaars, A. A. (2003). Can a theory of suicide predict all suicides in the elderly? Crisis, 24(1), 7-16. Lester, D. (1990). Current concepts of suicide. Philadelphia, Pennsylvania: The Charles Press, Publishers. Lewin, K. (1951). Constructs in field theory. In K. Lewin (D. Cartwright, Ed.), Field theory in social science: Selected theoretical papers (pp. 30-42). New York: Harper. (Originally published in 1944.) Lewin, K. (1952). Group decision and social change. In G. E. Swanson, T. M. Newcomb, & E. L. Hartley (Eds.), Readings in social psychology (2nd ed., pp. 459-473). New York: Holt, Rinehart, and Winston. Lewin, K., Lippett, R., & White, R. K. (1939). Patterns of aggressive behavior in experimentally created “social climates.” Journal of Social Psychology, 10, 271-299. Lippett, R., Polansky, N., Redl, F., & Rosen, S. (1952). The dynamics of power. Human Relations, 5, 37-64. 197 Lipschitz, A. (1990). College suicide: A review monograph. New York: American Suicide Foundation. Litman, R. E. (1967). Sigmund Freud on suicide. In E. Shneidman (Ed.), Essays in self-destruction. New York: Science House. Magoon, T. (2000). College and university counseling center directors’ data bank. Unpublished data. Maris, R. W. (1981). Pathways to suicide: A survey of self-destructive behaviors. Baltimore: Johns Hopkins University Press. Maris, R. W., Berman, A. L., Maltsberger, J. T., & Yufit, R. I. (1992). Assessment and Prediction of Suicide. New York: The Guilford Press. Maris, R. W., Berman, A. L., & Silverman, M. M. (2000). Comprehensive textbook of Suicidology. New York: The Guilford Press. McLaughlin, C. (2007). Suicide-related behavior: Understanding , caring, and therapeutic responses. West Sussex, England: John Wiley & Sons, Ltd. Meilman, P. W., Pattis, J. A., & Kraus-Zeilmann, D. (1994). Suicide attempts and threats on one college campus: Policy and practice. Journal of American College Health, 42(4), 147-154. Menninger, K. (1938). Man against himself. New York: Harcourt, Brace & World. Mintzberg, H. (1983). Power in and around organizations. Englewood Cliffs, NJ: Prentice Hall. 198 Mofidi, N., Ghazinour, M., Salander-Renberg, E., & Richter, J. (2008). Attitudes toward suicide among Kurdish people in Iran. Social Psychiatry and Psychiatric Epidemiology, 43, 291-298. Morgan, H., Evans, M., Johnson, C., & Stanton, R. (1996). Can a lecture influence attitudes to suicide prevention? Journal of the Royal Society of Medicine, 89(2), 87-90. Murray, H. (1938). Exploration in personality. New York: Oxford University Press. Murray, H. (1967). Death to the world: The passions of Herman Melville. In E. Shneidman (Ed.), Essays in self-destruction (pp. 7-29). New York: Science House. National Association of Student Personnel Administrators. About us. (n.d.). Retrieved August 17, 2009, from http://www.naspa.org/about/default.cfm. Nisbet, P. A. (1995). Suicidal protective factors in black females. Unpublished master’s thesis, University of South Carolina (Columbia). Nisbet, P. A. (1998). Interactive wealth and the opportunity for self-destructive behaviors. Unpublished doctoral dissertation, University of South Carolina (Columbia). Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. Nuss, E. M. (2003). The Development of Student Affairs. In S. R. Komives, D. B. Woodard Jr., and Associates (eds.), Student Services: A Handbook for the Profession. (4th ed.). San Francisco: Jossey-Bass. 199 Pescosolido, B. A., & Georgianna, S. (1989). Durkheim, suicide, and religion: Toward a network theory of suicide. American Sociological Review, 54, 33– 48. Pfeffer, J. (1981). Power in Organizations. Marshfield, MA: Pitman. Pickering, J. W., & Calliotte, J. A. (2000). Who are we? Where do we want to go? A report on the membership of the American College Personnel Association. Retrieved October 28, 2009, from http://www.acpa.nche.edu. Pickering, W. S. F. (Ed.). (2002). Durkheim Today. Berghahn Books: New York. Pierro, A., Cicero, L., & Raven, B. H. (2008). Motivated compliance with bases of social power. Journal of Applied Social Psychology, 38(7), 1921-1944. Platt, S. (1989). A subculture of parasuicide? In R. F. W. Diekstra, S. R. Maris, S. Platt, A. Schmidtke, & G. Sonnek (Eds.), Suicide and its prevention: The role of attitude and imitation (pp. 108-143). Leiden: Canberrs, Brill. Rahim, M. A. (1988). The development of a leader power inventory. Multivariate Behavioral Research, 23, 491-503. Raphael, T., Power, S. H., & Berridge, W. L. (1937). The question of suicide as a problem in college mental hygiene. American Journal of Orthopsychiatry, 7, 114. Raven, B. H. (1965). Social influence and power. In I. D. Steiner & M. Fishbein (eds.), Current studies in social psychology (pp. 371-381). New York: Holt, Rinehart & Winston. 200 Raven, B. H. (1993). The bases of power: Origins and recent developments. Journal of Social Issues, 49, 227-251. Rickard, S. T. (1985). Titles of chief student affairs officer: Institutional autonomy or professional standardization? NASPA Journal, 23(2), 44 – 49. Rogers, J. R., & DeShon, R. P. (1995). Cross-validation of the five factor interpretive model of suicide opinion questionnaire. Suicide and Life- Threatening Behavior, 25, 303-309. Rudd, M. D. (1989). The prevalence of suicidal ideation among college students. Suicide and Life-Threatening Behavior, 19, 173-183. Salander-Renberg, E., Hjelmeland, H., Koposov, R. (2008). Building models for the relationship between attitudes toward suicide and suicidal behaviors: Based on data from general population surveys in Sweden, Norway, and Russia. Suicide and Life-Threatening Behavior, 38(6), 661 – 675. Salander-Renberg, E., & Jacobsson, L. (2003). Development of a questionnaire on attitudes towards suicide (ATTS) and its application in a Swedish population. Suicide and Life-Threatening Behavior, 33, 52-64. Sale, I., Williams, C. L., Clark, J., & Mills, J. (1975). Suicide behavior: Community attitudes and beliefs. Suicide, 5, 158-168. Salkind, N. J. (2004). Statistics for people who (think they) hate statistics. (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc. Sandeen, A. (1991). The Chief Student Affairs Officer. San Francisco: Jossey- Bass. 201 Sandeen, A., & Barr, M. (2006). Critical issues for student affairs: Challenges and opportunities. San Francisco: Jossey-Bass. Scelfo, J. (2003, November 3). Preventing suicide. Newsweek, 142 (18), 4. Schwartz, A. J. (1995). Suicide rates, selectivity of admissions and academic competitiveness. Proceedings of the Annual Conference of the Association of University and College Counseling Center Directors, 2-14. Schwartz, A. J. (2006). College student suicide in the United States: 1990-1991 through 2003-2004. Journal of American College Health, 54(6), 341-352. Schwartz, A. J., & Reifler, C. B. (1988). College student suicide in the United States: Incidence data and prospects for demonstrating efficacy of preventative programs. Journal of American College Health, 37, 53-59. Schwartz, R. A. (1997). Reconceptualizing the leadership roles of women in higher education: A brief history on the importance of Deans of Women. The Journal of Higher Education, 68(5), 502 – 522. Schwartz, R. A. (2002). The rise and demise of Deans of Men. The Review of Higher Education, 26(2), 217 – 239. Secord, P. F., & Backman, C. W. (1974). Social Psychology. Tokyo: McGraw- Hill. Sheehan, K. (2001). Email survey response rates: A review 1986-2000. Journal of Computer-Mediated Communication, 6(2), 24-61. Shneidman, E. (1985). Definition of suicide. New York: Wiley. Shneidman, E. (1993). Psychache. Northvale, NJ: Aronson. 202 Silverman, M. M. (1993). Campus student suicide rates: Fact or artifact? Suicide & Life-Threatening Behavior, 23, 329-342. Silverman, M. M., Meyer, P. M., Sloane, F., Raffel, M., & Pratt, D. M. (1997). The Big Ten student suicide study: A 10-year study of suicides on Midwestern college campuses. Suicide and Life-Threatening Behavior, 27(3), 285-303. Stack, S. (1983). The effect of religious commitment on suicide. Journal of Health and Social Behavior, 24, 362-374. Stack, S. (2001). Sociological research into suicide. In D. Lester (Eds.), Suicide Prevention: Resources for the Millennium (pp. 17 -30). Stark, R., Doyle, D. P., & Rushing, L. (1983). Beyond Durkheim: Religion and suicide. Journal for the Scientific Study of Religion, 22, 120-131. Stogdill, R. M. (1969). Validity of leader behavior descriptions. Personnel Psychology, 22(7), 153-158. Student, K. R. (1968). Supervisory influence and work-group performance. Journal of Applied Psychology, 52, 188-194. Suicide Prevention Resource Center. (2004). Promoting mental health and preventing suicide in college and university settings. Newton, MA: Education Development Center, Inc. Sullivan, H. (1962). Schizophrenia as a human process. In H. Perry, N. Gorvell, and M. Gibbons (Eds.), The collected works of Harry Stack Sullivan (Vol. 2). New York: W. W. Norton. 203 Sullivan, H. (1964). The fusion of psychiatry and social science. In H. Perry, N. Gorvell, and M.Gibbons (Eds.), The collected works of Harry Stack Sullivan (Vol. 2). New York: W. W. Norton. Thomas, K. W. (1985). Power Base Inventory. Sterling Forest, Tuxedo, NY: XICOM. Thomas, K. W., & Thomas, G. F. (1991). Power Base Inventory. (2nd ed.). Woods Road, Tuxedo, NY: XICOM. Tull, A., & Freeman, J. P. (2008). Chief student affairs officer titles: Standardization of titles and broadening of labels. NASPA Journal, 45(2), 265 – 281. Turner, F. R. (1936). Report of the committee on the preparation for work as a dean of men. Yearbook of the National Association of the Deans of Women, 1936, 104 – 106. U. S. Department of Education, National Center for Education Statistics. (2001). Digest of Educational Statistics: Postsecondary Education. Retrieved April 11, 2008, from http://www.nces.ed.gov. U. S. Department of Education, National Center for Education Statistics. (2008). Digest of Education Statistics, 2007 (NCES 2008-022), Chapter 3. Warren, D. I. (1968). Power, visibility, and conformity in formal organizations. American Sociological Review, 33, 951-970. Webb, N. B. (1986). Before and after suicide: A preventative outreach program for colleges. Suicide and Life Threatening Behavior, 16, 469-480. 204 Westefeld, J. S., Homaifar, B., Spotts, J., Furr, S., Range, L., & Werth, J. L. (2005). Perceptions concerning college student suicide: Data from four universities. Suicide & Life-Threatening Behavior, 35, 640-645. Westefeld, J. S., & Pattillo, C. M. (1987). College students’ suicide: The case for national clearinghouse. Journal of College Student Personnel, 28, 34-38. Whitaker, L. C. (1986). Psychotherapy as opportunity to prevent college student suicide. Journal of College Student Psychotherapy, 1, 71-88. Williams, M. L., & Podsakoff, P. M. (1988). A meta-analysis of attitudinal and behavioral correlates of leader reward and punishment behaviors. Southern Management Association Proceedings, 161-163. Wrong, D. H. (1980). Power: Its forms, bases, and uses. New York: Harper & Row. Yukl, G., & Falbe, C. M. (1991). The importance of different power sources in downward and lateral relations. Journal of Applied Psychology, 77, 525 – 535. Zander, P. G., & Cohen, A. R. (1955). Attributed social power and group acceptance: A classroom experimental demonstration. Journal of Abnormal and Social Psychology, 51, 490-492. Zilboorg, G. (1936). Suicide among civilized and primitive races. American Journal of Psychiatry, 92, 1347-1369. APPENDIX A: ATTITUDES TOWARD SUICIDE QUESITONNAIRE (ATTS) 206 207 208 209 210 211 212 213 214 APPENDIX B: POWER BASE INVENTORY (PBI) 216 217 APPENDIX C: NASPA MEMBERSHIP LIST REQUEST POLICY FOR RESEARCH PURPOSES APPENDIX D: NASPA MEMBERSHIP LIST REQUEST FORM 220 221 222 APPENDIX E: CCP, INC. REQUEST FOR RESEARCH SUPPORT 224 225 226 APPENDIX F: SIGNED CCP INC. CONTRACT 228 APPENDIX G: INSTITUTIONAL REVIEW BOARD APPROVAL LETTER APPENDIX H: INVITATION TO PARTICIPATE IN RESEARCH APPENDIX I: MODIFIED SURVEY FORMAT OF ATTS AND PBI 232 233 234