<%BANNER%>

Jail Staff Professional Orientation and Attitudes towards Victimization

Permanent Link: http://ufdc.ufl.edu/UFE0024396/00001

Material Information

Title: Jail Staff Professional Orientation and Attitudes towards Victimization
Physical Description: 1 online resource (251 p.)
Language: english
Creator: Cook, Carrie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Criminology, Law and Society -- Dissertations, Academic -- UF
Genre: Criminology, Law, and Society thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The incidence of sexual assault in correctional institutions is a topic that has been considered with great apprehension. Most research has been conducted over the last 25 years and addresses the issue from both inmate and prison staff perspectives. Unfortunately, law and correctional policy is only recently addressing the issue of sexual assault in correctional institutions, and has reached little consensus about the nature and prevalence of the problem. In addition, there is a paucity of research on how correctional staff view, respond to, and prevent the incidence of inmate-on-inmate sexual assault and victimization in prisons. The research that does address the nature of this problem in state prisons suggests that correctional staff believe that inmate-on-inmate sexual assault is not rare, and between 23-46% believe inmates who have previously consented to sexual acts deserve rape (Eigenberg, 1989). The effects of experiencing sexual assault in correctional institutions are serious, and include fear, anxiety, continued assault, suicide, and depression (Bowker, 1980, 1982; Lockwood, 1980). The topic has increased salience following the 2003 passage of the Prison Rape Elimination Act (PREA), which seeks to address the issue of sexual assault in prisons. The purposes of the law are to develop national standards to prevent, detect, and punish prison sexual assault; to increase data on its incidence; and to increase the accountability of prison officials (S. Res. 1435, 2003). It is important to note, however, that efforts at data collection regarding the incidence of sexual assault in prison are focused at a state prison level. To my knowledge, there is no literature that addresses how staff in neglected research realms, specifically jails, identify and manage the problem of inmate-on-inmate sexual assault and victimization. The purpose of this research is to conduct research in this area that has been traditionally ignored by criminologists and sociologists.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Carrie Cook.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Lane, Jodi S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024396:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024396/00001

Material Information

Title: Jail Staff Professional Orientation and Attitudes towards Victimization
Physical Description: 1 online resource (251 p.)
Language: english
Creator: Cook, Carrie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Criminology, Law and Society -- Dissertations, Academic -- UF
Genre: Criminology, Law, and Society thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The incidence of sexual assault in correctional institutions is a topic that has been considered with great apprehension. Most research has been conducted over the last 25 years and addresses the issue from both inmate and prison staff perspectives. Unfortunately, law and correctional policy is only recently addressing the issue of sexual assault in correctional institutions, and has reached little consensus about the nature and prevalence of the problem. In addition, there is a paucity of research on how correctional staff view, respond to, and prevent the incidence of inmate-on-inmate sexual assault and victimization in prisons. The research that does address the nature of this problem in state prisons suggests that correctional staff believe that inmate-on-inmate sexual assault is not rare, and between 23-46% believe inmates who have previously consented to sexual acts deserve rape (Eigenberg, 1989). The effects of experiencing sexual assault in correctional institutions are serious, and include fear, anxiety, continued assault, suicide, and depression (Bowker, 1980, 1982; Lockwood, 1980). The topic has increased salience following the 2003 passage of the Prison Rape Elimination Act (PREA), which seeks to address the issue of sexual assault in prisons. The purposes of the law are to develop national standards to prevent, detect, and punish prison sexual assault; to increase data on its incidence; and to increase the accountability of prison officials (S. Res. 1435, 2003). It is important to note, however, that efforts at data collection regarding the incidence of sexual assault in prison are focused at a state prison level. To my knowledge, there is no literature that addresses how staff in neglected research realms, specifically jails, identify and manage the problem of inmate-on-inmate sexual assault and victimization. The purpose of this research is to conduct research in this area that has been traditionally ignored by criminologists and sociologists.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Carrie Cook.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Lane, Jodi S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024396:00001


This item has the following downloads:


Full Text

PAGE 1

JAIL STAFF PROFESSIONAL ORIENTATION & ATTITUDES TOWARDS VICTIMIZATION By CARRIE LYNN COOK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Carrie Lynn Cook 2

PAGE 3

To Forrest, my parents, and my family 3

PAGE 4

ACKNOWLEDGMENTS I would like to thank Jodi Lane for her grac ious gift of time and energy. She taught me much about scholarship and academia. I canno t remember a time when she was unable to consult with me about anything I needed, and for that I will alwa ys be grateful. I would be negligent in failing to acknowledge Lonn Lanza-Ka duce for challenging me in Law and Society. It was this class and a conversati on during office hours that influenced my interest in this topic. I am also extremely appreciative to Marian Bor g, Charles Frazier, and Joseph Spillane for their time and thoughtful suggestions regarding my re search. The contributions of each committee member improved the significance and quality of th is project. I also thank Saundra Westervelt and Agnes Baro for inspiring me through their selfless dedication to edu cation and students. I hope to emulate their passion in my own career. I thank my friends and colleagues employed by the Georgia Department of Corrections for their faith and encouragement during my tenure as a counselor. I would also like to thank the staff at participating jails for graciously accomm odating me during my research trips. I am honored to have met such dedicated and professional officers and administrators. I am forever grateful to my parents for their steady love and encouragement. They provided me with a joyful and happy childhood a nd filled my life with both adventure and stability. They were always by my side to help me realize my ambitions, often putting aside their own. My mother is my idol and my father is my hero. I also thank the other members of my family for their love, support, and understanding. My brothers and their wives are four of the coolest people I know; all are motivated and brillia nt. They, as well as my precious nieces and nephews, have always understood my academic ambitions and limited visits, and have continued to support me regardless. I would also like to th ank my grandparents for their love and guidance; their dedication to family is unparalleled. I am grateful to my frie nds and graduate school 4

PAGE 5

colleagues for their loyalty and friendship. I also thank the Cooks for providing laughter and love while I finished graduate school. Finally, I thank Forrest for his friendship and unconditional love and support. He has always been a fan of my dreams, and has not only stood by me as I accomplished them, but cheered for me every step of the way. He is th e most unselfish person I have ever known, and I will always admire his integrity, courage, spont aneity, and never-ending sense of calm. On many days, he is my oxygen, every day he is my life. 5

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES .........................................................................................................................10ABSTRACT ...................................................................................................................... .............12 CHAPTER 1 INTRODUCTION ................................................................................................................ ..14Introduction .................................................................................................................. ...........14Present Study ..........................................................................................................................18Specific Aims ..........................................................................................................................18Significance .................................................................................................................. ..........19Limitations and Parameters of the Study ................................................................................202 REVIEW OF PRIOR RESEARCH ........................................................................................22Introduction .................................................................................................................. ...........22Sexual Assault and Rape in Correctional Institutions ............................................................23Sexual Assault and Rape in Prison ..................................................................................24Victims and Perpetrators .................................................................................................26Dominance and Stigma ....................................................................................................27Effects of Sexual Assault .................................................................................................27Inmate Reporting and In stitutional Response ..................................................................29Victim Bl aming ................................................................................................................ ......30Female Rape Victims ......................................................................................................30Male Rape Victims ..........................................................................................................33Comparisons of Blaming for Female and Male Rape Victims ........................................35Homosexual Rape Victims ..............................................................................................36Victim Blaming by Inmates .............................................................................................38Credibility of Rape Victims ................................................................................................... .39Definitions of Rape ........................................................................................................... ......40Willingness to Respond ..........................................................................................................41Professional Orientation of Correctional Officers ..................................................................42Counseling Roles, Rehabilitation, and Punitiveness .......................................................42Concern with Corruption of Authority ............................................................................44Social Distance ................................................................................................................44Jail Culture and Context ...................................................................................................... ...45Conclusion .................................................................................................................... ..........50 6

PAGE 7

3 DATA AND METHODOLOGY ...........................................................................................52Introduction .................................................................................................................. ...........52Research Questions ............................................................................................................ .....53Settings ...................................................................................................................... .............54Solicitation of Sites ......................................................................................................... .54Facility Participation .......................................................................................................55Survey Methodology ..............................................................................................................56Jail Characteristics ..................................................................................................................58Sample ........................................................................................................................ ............60Data .......................................................................................................................... ...............66Officer Perceptions of Victims ........................................................................................66Independent variables ...............................................................................................66Dependent variables .................................................................................................70Perceptions of Fellow Officers Professional Orientation ...............................................73Perceptions of Inmate Attitudes ......................................................................................74Analysis ..................................................................................................................................74Officer Perceptions of Victims ........................................................................................74Perceptions of Professional Orie ntation and Inmate Attitudes ........................................784 DESCRIPTIVE STATISTICS ................................................................................................88Introduction .................................................................................................................. ...........88Independent Variables ......................................................................................................... ...88Professional Orientation ..................................................................................................89Attitudes toward Homosexuality .....................................................................................93Male Rape Myths ............................................................................................................94Dependent Variables ........................................................................................................... ....95Victim Bl aming ...............................................................................................................9 6Inmate Credibility ............................................................................................................ 96Rape Definitions ..............................................................................................................97Willingness to Respond ...................................................................................................98Conclusion .................................................................................................................... ..........995 PERCEPTIONS OF VICTIMS QUANTITATIVE RESULTS ...........................................114Introduction .................................................................................................................. .........114Victim Bl aming ................................................................................................................ ....114Blaming and Inmate Characteristics ..............................................................................115Bivariate Correlations ....................................................................................................116Multivariate Analysis ....................................................................................................119Summary of Victim Blaming Analysis .........................................................................122Inmate Credibility ............................................................................................................ .....122Credibility and Inmate Characteristics ..........................................................................123Bivariate Correlations ....................................................................................................126Multivariate Analysis ....................................................................................................127 7

PAGE 8

Full sample .............................................................................................................128Split samples ..........................................................................................................129Summary of Inmate Credibility Analysis ......................................................................132Conclusion .................................................................................................................... ........1336 DEFINITIONS AND WILLINGNESS TO RESPOND QUANTITATIVE RESULTS .....145Introduction .................................................................................................................. .........145Rape Definitions ...................................................................................................................145Definitions and Context .................................................................................................146Bivariate Correlations ....................................................................................................148Multivariate Analysis ....................................................................................................150Summary of Rape Definition Analysis ..........................................................................151Willingness to Respond ........................................................................................................151Differences in Response Type .......................................................................................152Responding to consensual sexual acts ....................................................................152Responding to sexual assault ..................................................................................153Comparison of responses for cons ensual sex and sexual assault ...........................155Bivariate Correlations ....................................................................................................156Multivariate Analysis ....................................................................................................158Encouraging inmate reporting and prevention-full sample ....................................159Talking to inmates ..................................................................................................166Proactive measures .................................................................................................167Summary of Willingness to Respond Analysis .............................................................169Conclusion .................................................................................................................... ........1727 PERCEPTIONS OF PROFESSIONAL OR IENTATION AND INMATE ATTITUDES QUATITATIVE RESULTS .................................................................................................185Introduction .................................................................................................................. .........185Perceptions of Fellow Officers Professional Orientation ....................................................185Counseling Roles ...........................................................................................................186Punitive Orientation .......................................................................................................187Corruption of Authority .................................................................................................188Social Distance ..............................................................................................................190Perceptions of Inmate Attitudes ............................................................................................191Victim Bl aming .............................................................................................................191Male Rape Myths ..........................................................................................................193Conclusion .................................................................................................................... ........1948 CONCLUSION AND DISCUSSION ..................................................................................198Introduction .................................................................................................................. .........198Lessons ....................................................................................................................... ..........198Victim Bl aming .............................................................................................................199Inmate Credibility ..........................................................................................................20 1Rape Definitions ............................................................................................................205 8

PAGE 9

Willingness to Respond .................................................................................................206Encouraging inmate reporting and prevention .......................................................207Talking to inmates ..................................................................................................209Proactive measures .................................................................................................209Perceptions of Fellow Officers Professional Orientation .............................................211Perceptions of Inmate Attitudes ....................................................................................211Implications .................................................................................................................. ........212Practical Implications ....................................................................................................212Theoretical Implications ................................................................................................216Limitations and Future Research ..........................................................................................218Concluding Thoughts ........................................................................................................... .221APPENDIX: SURVEY INSTRUMENT .....................................................................................223LIST OF REFERENCES .............................................................................................................232BIOGRAPHICAL SKETCH .......................................................................................................251 9

PAGE 10

LIST OF TABLES Table page 3-1 Population, sample, and response rate by facility ..............................................................803-2 Facility characteristics .................................................................................................. .....803-3 Jail and sample characteristics ........................................................................................... 813-4 Factor loadings and reliability scores of indepe ndent variables ........................................843-5 Factor loadings and reliability scores of dependent variables ...........................................864-1 Descriptive statistics by sex: independent variables ........................................................1024-2 Descriptive statistics by edu cation: independent variables ..............................................1054-3 Descriptive statistics by sex: dependent variables ...........................................................1084-4 Descriptive statistics by e ducation: dependent variables .................................................1115-1 Victim blaming paired sample t -tests ..............................................................................1365-2 Victim blaming zero-order correlations ...........................................................................1385-3 Stepwise linear regression predicting victim blaming .....................................................1395-4 Inmate credibility paired sample t -tests ...........................................................................1405-5 Inmate credibility ze ro-order correlations .......................................................................1425-6 Stepwise linear regression predicting inmate credibility .................................................1436-1 Rape definitions paired sample t -tests .............................................................................1756-2 Definitions of rape zero-order correlations ......................................................................1766-3 Stepwise linear regression predicting definitions of rape ................................................1776-4 Willingness to respond paired sample t -tests ...................................................................1786-5 Willingness to respond zero-order correlations ...............................................................1806-6 Stepwise linear regression predicting willingness to respond by encouraging inmate reporting & officer prevention .........................................................................................1816-7 Stepwise linear regression predicting w illingness to respond by talking to inmates .......183 10

PAGE 11

6-8 Stepwise linear regression predic ting willingness to respond proactively ......................1847-1 Perceptions of fellow officer prof essional orientation paired sample t -tests ...................1957-2 Perception of inmate attitudes paired sample t -tests ........................................................197 11

PAGE 12

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy JAIL STAFF PROFESSIONAL ORIENTATION & ATTITUDES TOWARDS VICTIMIZATION By Carrie Lynn Cook May 2009 Chair: Jodi Lane Major: Criminology, Law, and Society The incidence of sexual assault in correcti onal institutions is a topic that has been considered with great apprehensi on. Most research has been conducted over the last 25 years and addresses the issue from both inmate and pris on staff perspectives. Unfortunately, law and correctional policy is only recen tly addressing the issue of se xual assault in correctional institutions, and has reached li ttle consensus about the nature and prevalence of the problem. In addition, there is a paucity of research on how correctional staff view, respond to, and prevent the incidence of inmate-on-inmate sexua l assault and victimization in prisons. The research that does address the natu re of this problem in state pr isons suggests that correctional staff believe that inmate-on-inmate sexual as sault is not rare, and between 23-46% believe inmates who have previously consented to se xual acts deserve rape (Eigenberg, 1989). The effects of experiencing sexual assa ult in correctional institutions are serious, and include fear, anxiety, continued assault, suicide, and depression (Bowker, 1980, 1982; Lockwood, 1980). The topic has increased salience followi ng the 2003 passage of the Prison Rape Elimination Act (PREA), which seeks to address the issue of sexual assault in prisons. The purposes of the law are to develop national standa rds to prevent, detect, and punish prison sexual assault; to increase data on its incidence; and to increase the accountability of prison officials (S. 12

PAGE 13

13 Res. 1435, 2003). It is important to note, however, that efforts at data collection regarding the incidence of sexual assault in pr ison are focused at a state prison level. To my knowledge, there is no literature that addresses how staff in neglected research realms, specifically jails, identify and manage the problem of inma te-on-inmate sexual assault and victimization. The purpose of this research is to conduct research in this area that ha s been traditionally ignored by criminologists and sociologists.

PAGE 14

CHAPTER 1 INTRODUCTION Introduction The incidence of sexual assault in correcti onal institutions is a topic that has been considered with great apprehension. Most research conducted over the last 25 years addresses the issue from both inmate and prison staff pers pectives. Unfortunately, law and correctional policy is only recently addressing th e issue of sexual assault in co rrectional institutions, and has reached little consensus about the nature and pr evalence of the problem. Only recently has the judicial system declared sexual assault in prison as cruel and u nusual punishment, a violation of the 8th amendment ( Farmer v. Brennan, 1994). In addition, there is a paucity of research on how correctional staff view, respond to, and prevent the incidence of inmate-to-inmate se xual assault and victimiz ation in correctional institutions. The research that does address the nature of this problem in state prisons suggests that correctional staff believe that inmate-to-in mate sexual assault is a common occurrence and that they believe that consensual homosexuality is less common than rape (Eigenberg, 1989). However, another Eigenberg study (2000a) found th at officers estimated that about 18% of inmates had been victims of rape, while 25% partic ipated in consensual acts of homosexuality. Recent research using surveys of inmates typica lly reach similar conclusions and estimate that around 20% of inmates report bei ng pressured or coerced into sex (Hensley, Koscheski, & Tewksbury, 2005; Lockwood, 1980; Struckman-Johnson,Struckman-Johnson,Rucker, Bumby, & Donaldson, 1996; Struckman-Johnson & Struckman-Johnson, 2000). Yet research suggests that correctional officers admit having trouble distinguishing between consensual and forced sexual acts between inmates (Eigenberg, 2000b; National Institute of Corrections & The Moss Group, In c., 2006). In addition, between 12-46% of the 14

PAGE 15

officers in one study believed inmates who have previously been ra ped deserve continued victimization (Eigenberg, 1989; Ei genberg, 2000a). Most research on victims of rape and sexual assault in the community also s uggests that there is a significant amount of victim blaming that attributes responsibility of the in cident and a lack of credibility to the victim. In addition, it appears that victim blaming may be more significant overall for certa in victims of sexual assault. Specifically, victims of date rape, or who have had prior intimacy with their assailant are seen as responsible (Bell, Kuriloff, & Lottes, 1994; Schuller & Hasti ngs, 2002). Victims who have had previous sexual experiences or sexual as saults are also seen as more responsible for their victimization (Pugh, 1983). Victims who do not resist or report immediately are perceived as atypical (not genuine victims) and more culpable (Calhoun, Selby, &Warring, 1976; Krulewitz & Nash, 1980; Spears & Spohn, 1996). Perceiving victims as provocative also increases perceptions of culpability (Kanekar & Kolsawalla, 1980). Finally, victims perceived to have experience pleasure or those who should have foreseen the attack (McCaul, Veltum, Boyechko, & Crawford, 1990). The limited research on victim blaming among correctional officers suggests that opposition to homosexuality increases the probabili ty of blaming victims (Eigenberg 2000a). However, there is no other research that sheds light on whether correctional officers perceive some inmates as more blameworthy and less credib le than other inmates. Specifically, there is no research that I am aware of that addresses jail correctional o fficer perceptions regarding this issue. The psychological effects for inmates who experience sexual assa ult in correctional institutions are serious, and can include: fear, anxiety, suicide, depression continued assault, anger, loss of identity, manhood, and self-esteem, increased distrust of others, disruption in 15

PAGE 16

social relationships, withdrawal, isolation, in creased aggressiveness and violence, negative sexual attitudes, and continued assau lt ( (Bowker, 1980, 1982; Cotton & Groth, 1982; Lockwood, 1980). These effects can manifest in further problems for institutional management and inmate adjustment; however, the consequences of these psychological scars can also be dire for the communities and families to which inmates return. In addition, the incidence of sexual assault can cause medical problems for inmates and facilitate the spread of infectious disease both inside prison and in the communities to which they are released. Ther e is also an increased need for medical services inside institutions for inmates that become sexually victimized and report their victimization or seek medical care. (Corlew, 2006; Lockwood, 1980). Instances of sexual assault and rape in institutions can breed additional acts of assault and violence (Corlew, 2006; Cotton & Groth, 1982; Fish man, 1934) and therefore pose grave risk to inmates and officers as well as significant institutional management problems. Additional acts of violence that may occur serve either retaliatory or protection purposes, esp ecially if inmates feel that correctional officers are not willing or able to protect them (Corlew, 2006; Fishman, 1934). Another often overlooked consequen ce of failing to address instances of sexual assault and rape in correctional institutions is the undermini ng of correctional authority and management. Correctional authorities that are aware of and fa il to address sexual assault in their institutions may find that inmates in their faci lity doubt the level of control that authorities have over their institution (Corlew, 2006). This, in turn, could lead to increase d disciplinary and management problems, as well as increased violence. Contro lling and preventing any act of violence within correctional institutions takes resources, includi ng additional fiscal and personnel resources, and can often strain already li mited correctional budgets. 16

PAGE 17

The topic of sexual assault in correctional institutions has increased salience following the 2003 passage of the Prison Rape Elimination Act (PREA), which seeks to address the issue of sexual assault in prisons. The purposes of the law are to devel op national standards to prevent, detect, and punish prison sexual assaul t; to increase data on its incidence; and to increase the accountability of pr ison officials (S. Res. 1435, 2003). One of the requirements of PREA is for the Bure au of Justice Statistics to collect data on the incidence of sexual violence through inmate se lf-reports and analysis of official institutional records (Beck, Harrison, & Adams, 2007; S. Re s. 1435, 2003). When data collection began in 2004, correctional authorities re ported 2.46 allegations of sexual violence per 1,000 adult inmates in prison, jails, and other correctiona l facilities. This figure rose to 2.91 in 2006, although this increase may be the result of data measurement techniques (Beck, et al., 2007). Data collection efforts to interview inmate s and victims of sexual assault through Audio Computer-Assisted Self-Inter view (ACASI) methodology began in 2006 and are expected to improve estimates of self-reported victimiza tion (Beck et al., 2007; Beck & Hughes, 2005). Recent results from this study indicate that about 3.2% of sampled jail inmates experienced sexual victimization, 1.6% experienced this victimization by other inmates (Beck & Harrison, 2008). The importance of data collection efforts ai med at understanding the incidence, nature, and culture of sexual assault in correctional in stitutions from the perspective of inmates and victims cannot be disputed, however, current research efforts do not include a study of correctional officer attitudes and response towards this issue. It is also important to note that most efforts at data collection re garding the incidence of sexual assault in prison are focused at the state prison level. 17

PAGE 18

Present Study There is limited research on how correctional staff view and respond to the incidence of inmate-to-inmate sexual assault and victimizatio n in prisons. To my knowledge, there is no literature that addresses how sta ff in neglected research realms specifically jails, view and respond to the problem of inmate-to-inmate sexual assault and victimization. The purpose of this study is to address shortcomings in research by examining an area that has been traditionally ignored by social scientists, specifically how correctional officers in jail think about and respond to this issue. This research will further the understanding of jail correctional officers professional orientations, attitudes toward homosexuality, and attitudes to ward sexual assault and rape victims in jail. Additionally, it will explore jail corre ctional officer percepti ons about the role of the victim in precipitation of incide nts of sexual assault and rape in jail, the credibility of victims, definitions of sexual assault and rape, and corr ectional officer willingness to respond to these incidents. Specific Aims One purpose of this research is to examine co rrectional officers attitudes regarding sexual assault, rape and homosexuality in jail. The spec ific research questions that I attempt to answer regarding perceptions of sexual assau lt and sexual assault victims include: 1. How do correctional officers perceive victims of sexual assault in jail? Specifically, do they attribute blame to victims of ra pe or sexual assault (victim-blame)? 2. How do officers perceive the credibility of vi ctims of sexual assault in jail? Are all inmates equally credible as victims of sexual a ssault, or are there ce rtain inmates that are more likely to be believed as victims? 3. How do officers define rape in a jail atmosphere? Specifica lly, what are their attitudes regarding sexual consent af ter coercion or threats? 18

PAGE 19

4. How do officers perceive their own willingness to respond to acts of sexual assault and rape in jail? Another purpose of the research is to examine correctional officer perceptions about the beliefs of their fellow officers a nd inmates in their facilities. The specific questions that this study explores regarding officer perceptions of th e attitudes of their fellow officers and inmates in include: 5. Are correctional officers in jails in Florid a accurate in their perceptions about the professional orientation of ot her officers? Specifically, how precisely do correctional officers identify their colleagues beliefs rega rding counseling roles, punitive orientation, corruption of authority, and soci al distance with inmates? 6. How do correctional officer perceive the attitudes of inmates in their facility regarding victim blaming and belief in male rape myth s? How do these perceptions compare to their own attitudes of victim blaming and male rape myth? Significance Although there is more research in this area, traditionally, the resear ch that does exist has shown that correctional officers do not believe that sexual assault in their institutions is rare (Eigenberg, 1989). Recent research on the incidence of sexual assa ult in institutions shows that allegations are as high as 2.91 per 1,000 inmates in adult correctional faci lities, including both prisons and jails (Beck et al., 2007). The contributions of this research are primar ily applied. Instances of sexual assault and rape in institutions can often breed additional acts of assault and violence in institutions (Corlew, 2006; Dynes & Donaldson, 1992; Lockwood, 1980) a nd therefore pose thr eats to both inmates and officers. In addition, there may also be collateral consequences for families of inmates who have experienced sexual violence during inca rceration (Corlew, 2006; Dynes & Donaldson, 1992; Lockwood, 1980). These incidences also cau se institutional management problems. Preventing any act of violence within correctio nal institutions takes resources, including additional fiscal and personnel resources, and ca n often strain alrea dy limited correctional 19

PAGE 20

budgets. Correctional officers in an y institution are forefront witn esses to these incidents, and are therefore key actors in the management of these occurrences. Understanding how they perceive inmates involved in this behavior and their willingness to intervene has implications for institutional management of inmate -to-inmate sexual assault and rape. The research will be useful in both furthe ring our understanding of correctional officers and the specific culture of jail correctional officers. The findings of this research will be interpreted in a victimology framework, specifica lly using victim blaming and the just world perspective as a discussion guide. The findings will inform further understanding of the work organization, subculture, and experiences of correctional officers in an over looked area, jails. In addition, the study will facilitate victimology research by exploring attributions of blame against a highly marginalized victim population. Limitations and Parameters of the Study One limitation of this research is its lack of external validity; the study was only conducted in one state and in facilities that chos e to participate. Generalizability is further limited to those correctional officers who chose to complete and return the survey. Because of these limitations, the findings may be an artifact of both the re search methodology and characteristics of participating ja ils. Another limitation of the re search is a low response rate (32.4%), which further limits generalizability. The findings could be a result of selection bias, or pre-existing characteristics of the participants who completed and returned the survey. The questions on the survey deal only with attitudes toward victimization between male inmates and do not address the issue of female-to-female inmate victimization. Self-report methodology relies on participant truthfulness in responding to questi ons, and the findings rest on the assumption that participants reported th eir perceptions and a ttitudes accurately. 20

PAGE 21

21 Regardless of these limitations, this research is important as it explor es a topic that has been given little atte ntion historically. Though the issu e of sexual assault and rape in correctional institutions is recen tly gaining attention in political and academic arenas, research about the views of correctional staff remain understudied. Ther e is very limited research on correctional officer attitudes toward the sexual victimization and inmates in general; the few studies that have been conducted in this area focus only on prison co rrectional officers. Research on professional orientations of correctional officers has also been conducted primarily in state or federal prison institutions. It is important to explore these topics among jail correctional staff as jail facilities function as the primary entrance into the system and annually process a larger population of inmates than state or federal prison facilities. This research contributes to an understanding of jail corre ctional officers professional orient ations and attitudes toward the sexual assault and rape of t hose whom they supervise.

PAGE 22

CHAPTER 2 REVIEW OF PRIOR RESEARCH Introduction The problem of sexual assault among inmates in correctional institutions is one that has only recently garnered attention from researchers and public officials. The perception of victims of sexual assault, though given more attention in research literature, has focused primarily on attitudes toward female victims of male perpetrated rape. The vast majority of this research explores perceptions of the public or college stud ents regarding rape victims, and uses vignettes to adjust characteristics of the offender, the vi ctim, and the situational context of the incident (Calhoun et al., 1976; FeldmanSummers & Lindner, 1976; Foley, Evancic, Karnick, King, & Parks, 1995; Kanekar & Kolsawalla, 1980; K opper, 1996; Krulewitz & Nash, 1980; MaCrae & Shepherd, 1989; Schuller & Hastings, 2002; Smith, Keating, Hester & Mitchell, 1976). Considering the many studies that examine attitudes about rape victims, only a small portion of the them address this topic in regard to male or homosexua l rape victims (Burt & DeMello, 2002; Davies, Pollard, & Archer, 2006; Doherty & Anderson, 2004; Ford, LiwagMcLamb & Foley, 1998; Hodge & Canter, 1998 ; King & Woollett, 1997; Masters, 1986; Mitchell, Hirschman, & Nagayama Hall, 1999; Perrott & Webber, 1996; Stermac, del Bove, & Addison, 2004; Stuart & Greer, 1984; White & Robins on Kurpius, 2002). This lack of attention may be because male victims compose such a small proportion of all rape victims (Tjaden & Thoennes, 2006). Research attempting to unders tand how people view sexual assault and rape victims has also traditionally ignored other marginalized victims, es pecially those who are incarcerated. Eigenberg conducted some of the first res earch focused at unde rstanding how those employed in correctional institutions view sexual assault and victimization among inmates 22

PAGE 23

(Eigenberg, 1989; 1994; 2000a; 2000b). Although her research is fo cused in prison settings, it represents the first effort to ex amine the previously ignored topi c of how correctional staff view victims of sexual assault and ra pe in correctional institutions No research has addressed attitudes of correctional officers in jails re garding incarcerated sexual assault victims. Jails, unlike prisons, have larg ely been ignored as a subject of research until relatively recently (Applegate & Sitren, 2008; Gibbs 1983; Klofas, 1990; Mays & Thompson, 1988; Welsh, 1992). Though there have been efforts focu sed at understanding the perceptions of jail personnel, these studies explore t opics such as job satis faction, stress, and use of force (Castle & Martin, 2006; Griffin, 2002; He mmens & Stohr, 2001; Pogrebin & Poole, 1998). Overall, research about the perceptions of jail correctiona l officers is scarce considering the prevalence of jails in the United States. This chapter offers an in-depth review of the existing research on rape in correctional institutions and staff attitudes toward inmates a nd victims of sexual assault in these institutions. The chapter continues with a revi ew of attitudes about rape victim s in general, and male rape victims specifically. I will also investigate how rape myths affect attitudes toward victims of rape. Finally, this chapter offers a review of the jail setting as th e context of the research by first identifying general features of jails and explor ing differences that may characterize jails across size and setting. Sexual Assault and Rape in Correctional Institutions Historically, the topic of sexual assault and rape in correctional institutions has been given very little attention. Res earch that has explored this issu e in the last several decades has typically focused on uncovering the prevalence and incidence of sexua l assault and rape in state prisons contexts generally. Even though the to pic has recently been explored in empirical research, there is a lack of consensus about the tr ue incidence and prevalen ce of sexual assault in 23

PAGE 24

correctional institutions. Some s uggest that the lack of clear unde rstanding of the nature of this topic is partially a function of prison, inmate a nd staff subcultures and attitudes (Cotton & Groth, 1982; Eigenberg, 1989). Inmates ar e hesitant to discuss what happens between inmates with officials because a convict code insists that they mind their own busin ess and not cooperate with authorities (Clear, Cole, & Reisig, 2006; Hassine, 2004; Johnson, 2002; Santos, 2004; Sykes, 1958). Officials may also be hesitant to pursue this topic for the same reasons that they are hesitant to discuss other sensitive issues like staff misconduct and corruption. They may not want to recognize its occurrence in their facilities for fear of appearing unable to manage effectively (Dennehy & Natel, 2006; Flannery, 1980; Murton, 1976). In theses ways, the culture of correctional institutions at tributes to the secrecy behind and ignorance of this issue. Sexual Assault and Rape in Prison Most studies of prison rape have attempted to uncover the incidence of sexual assault in prisons; however, one exception is discussed in more detail below. In a study of New York prisons, Lockwood (1980) found that 28% of the 76 inmates interviewed had been targeted by sexual aggressors, although only one inmate reported being a victim of rape. Similar findings were reported by Tewksbury (1989), while he found no inmate who self-reported being a victim of completed sexual assault, inmates did report be ing approached for sexual favors in a forceful manner. In Lockwoods (1980) analysis, half of the inci dents contained violence and a third of the targets suffered physical injuries. He also found that a quarter of all inmate-to-inmate physical assaults had sexual motivations at their core. Nacci and Kane ( 1983), in a study of federal prison inmates, found that a quarter of major assaults could be traced to homos exual activity. Five of eight homicides in one federal prison were cau sed by sexual motivations. In an exploratory analysis of prison rape, Fishman (1934: 70) observed that fights of a deadly nature occur 24

PAGE 25

between prison inmates who had fallen in love or were seeking the favor of one of these known homosexuals. These finding s indicate that the incidence of prison violence cannot be understood without attempting to understand the culture of sexual activity in prisons. Overall, Nacci and Kane (1983) found that 12 -30% of inmates reported participation in homosexual activities in prison, while 29% ha d been propositioned for homosexual activity. Other studies show much higher rates of victimization of comp leted sexual assault or rape. Struckman-Johnson et al. (1996), in a 1994 survey, found that 20% of inmates reported that they had been pressured or forced into sex. About 10% were coerced with tactics like bribes and blackmail, while 75% reported coercion through th reat of force or actual force (StruckmanJohnson et al., 1996). In another study administer ed in 1998, some researchers found that 21% of over 7,000 inmates self-reported pressured or for ced sexual activity duri ng their incarceration (Struckman-Johnson & Struckman-Johnson, 2000). Hensley et al. (2005) found that 18% of inmates in a maximum-security male facility had been targets of sexual propositions and about 8% were victims of sexual assault. In a study measuring inmate perceptions of prison rape, Saum, Surratt, Inciardi & Bennett (1995) found that half of the inmates surveyed in a Dela ware treatment program reported hearing about consensual sex acts between inmates. A quarter of these inmates reported that they had witnessed consensual sex between inmates. Re garding rape, 60% reported that they had not heard of any prison rapes, and only 4% reported witnessing rape between inmates. One inmate (of 101 interviewed) reported being raped while in prison, while five inmates reported being victims of rape attempts. In a study of inmates in a medium security California prison, Woode n and Parker (1982) found that 14% of inmates reported being victims of sexual assault. No t distinguishing between 25

PAGE 26

consensual and non-consensual acts, they found th at 65% of their sample had engaged in prison sex. One recent study of self-reports among ja il inmates nationwide s uggests that 3.2% of inmates experienced sexual victimization and 1.6% of inmates experienced this victimization by other inmates (Beck & Harrison, 208). Although th ese studies reflect different measures of incidence, they indicate that sexual assault in prison does occur and should be addressed. Victims and Perpetrators A substantial amount of resear ch attempts to classify ch aracteristics of victims and perpetrators of prison sexual assa ult and informs a broa der understanding of the culture of these incidents. Several studies identify the follo wing characteristics of potential victims: younger than perpetrators, physically weak or small, nervous in appearance, talkative, feminine or boyish, willing to accept favors, not prone to violence, ha ving a history of mental health treatment, not being streetwise, having convict ions for sexual offenses, and ha ving the reputation of a snitch (Bowker, 1980, 1982; Chonco, 1989; Dumond, 2000; Fishman, 1934; Lockwood, 1980; McFarland, Ellis, & Chunn, 2008; Weiss & Fria r, 1974; Wooden & Parker, 1982) Nonheterosexuals, compared to heterosexuals, and th ose who have been sexually victimized in the past are also likely to be victimized (B eck & Harrison, 2008; McFa rland et al., 2008). While some research indicates that victim s have no prior incar ceration experience (Bowker, 1980; Chonco, 1989; McFarland et al., 2008), one study shows that over half of targets of sexual assault self-report previous incarceratio ns (Lockwood, 1980). Victims are seen as passive and labeled in derogatory ways (e.g., fags queens, punks, sissies), while perpetrators are labeled with more aggressive terms (e.g., wolv es, jockers, studs) (Fishman, 1934; Wooden & Parker, 1982). Victims are more likely to be White, wh ile non-Whites are over-represented as perpetrators (Hensley et al., 2005; Lockwood, 1980; Tewk sbury, 1989; Wortley, 2002). 26

PAGE 27

Perpetrators are described as young (but generally older than victims) with bigger physical statures. They are also considered more confid ent, manipulative, and aggressive than victims are. Finally, perpetrators ofte n have prior offenses, prior inca rcerations, long sentences, and more serious offenses (Chonco, 1989; McFarland et al., 2008). Perpetrators sometimes test potential victims to see if they are willing to accept favors, which reflects weakness and dependence. It is through acceptance of favors or protection that perpetrators may set up the stage for quid pro quo victimization, where they coerce or threaten sexu al assault as payment for the previously accepted favor (Chonco, 1989; Fishman, 1934; Weiss & Friar, 1974). Dominance and Stigma Researchers have concluded that prison rape is the expression of dominance and power that does not carry the stigma within prison that it does outside the prison (ODonnell, 2004). For example, male rapists are seen neither as rapists, nor as homosexuals (ODonnell, 2004; Weiss & Friar, 1974). Taking on the role of the dominant partner, the aggressor is not viewed as participating in homosexuality by his inmate peers (Hassine, 2004; Wooden & Parker, 1982). Being an aggressive pursuer can enhance an inmate s status as a dominant male. It is when an inmate participates in the passive role of sex that he is viewed as homosexual. Prison sex is defined by its physical, rather than emotiona l relationship (Wooden & Parker, 1982) and prison rape is often accompanied by brutality and sadism (Bowker, 1980; Lockwood, 1980). Effects of Sexual Assault In one of the most graphic published account s of sexual victimization and rape in institutions, Tucker (1992) de scribes how he was forcefully raped upon entering a Washington D.C. jail. The main effect of rape for him wa s a transition into a lifelong institutional punk who became willing to provide sexual servic es for protection from others. His account suggests inmate intent to turn him into a prostitute for the utility of other inmates. The idea that 27

PAGE 28

recruitment into institutional homosexuality serv es the purposes of the prison culture is also reflected elsewhere in the litera ture (Van Wormer, 1984). Tucker (1992) implies that his desire to escape rape by many inmates led him to secure a protected relationship w ith one, and that this attempt to avoid emotional isolation is as cl ose to a voluntary re lationship as he came. Psychological effects of sexual assault in correctional instit utions, described by victims, include depression and attempted suicide, fear, anxiety, anger, increased distrust of others, disruption in social relationshi ps, withdrawal, isolation, incr eased aggressiveness, negative sexual attitudes, continued assault, and loss of identity, manhood, and self-esteem (Bowker, 1980, 1982; Cotton & Groth, 1982; Lockwood, 1980, Tucker, 1992). Some victims may even submit to further victimization due to fear and threats of physical assa ult and death (Tucker, 1992). Victims of sexual assault in prison report te ndencies for racism due to the inter-racial nature of assault, and the development of vi olent tendencies for protection (Bowker, 1980, 1982; Lockwood, 1980). Others become self-destructive and commit acts of self-harm, including suicide attempts (Bowker, 1980). Forced to commit homosexual activity against their will, victims often lose their identi ties as heterosexuals or real men (Bowker, 1980). StruckmanJohnson et al. (1996) found that most inmates re ported at least one emotional consequence of their experience (i.e., distrust, nervousness, depression), while 16% of victims experienced physical injuries from the incident. Sexual assault can also cause medical problems for inmates and facilitate the spread of infectious disease (Corlew, 2006; Cotton & Groth, 1982). Instances of sexua l assault and rape in institutions can breed additional acts of assau lt and violence and therefore pose grave risk to 28

PAGE 29

inmates and officers as well as significant inst itutional management problems (Cotton & Groth, 1982; Fishman, 1934; Nacci & Kane, 1983). Inmate Reporting and Institutional Response One of the commonly accepted tenets of this topic is that inmates underreport instances of sexual assault and rape in correctional instit utions. There are vari ous reasons why inmates may not report acts of sexual a ssault and rape, including the per ception that prison officials will not believe them, will not pursue prosecution, the fear of being labeled a snitch, the fear of further victimization, and the fear of being blamed for the incident (Dumond, 2000; Dumond & Dumond, 2002; Eigenberg, 1989; Fishman, 1934; Hassine, 2004; Santos, 2004; Wooden & Parker, 1982). Historically, suggestions to prevent sexual as sault and rape in correctional institutions have included segregation of potential and actual victimizer s and targets and increased surveillance of facility blind spot s where assaults are likely to occur (Cotton & Groth, 1982). Cotton & Groth (1982) also provide an overview of options av ailable to inmates who find themselves targets of sexual assault. They stipulate that inmates may request protective custody; however, they may subsequently give up certain liberties and privileges that accompany inmates in the general population (Cotton & Groth, 1982). Th ey also state that inmates may fight back and risk greater injury and disc iplinary action or they may submit to the sexual assaults (making them vulnerable for further victimization) (Cotton & Groth, 1982). These options have implications for the management of institutions; another implication for prison management is that males are hesitant to seek assistance in the first place and may be more prone to try to take care of the situation themselves (Cotton & Gr oth, 1982; Hassine, 2004). In addition, one of the main functions of prison rape in their study was retaliation (Cotton & Groth, 1982). 29

PAGE 30

Although correctional officers work in settings in which inmate culture can influence beliefs about prison rape, the resu lts of several studies indicate that there is misunderstanding among correctional personnel about th e nature and effects of inmate sexual assaults (Eigenberg, 2000b). Though prison rape is thought to genera te many institutional problems (Corlew, 2006; Dynes & Donaldson, 1992; Fishman, 1934; Lockw ood, 1980), there is a paucity of research regarding correctional staff attitudes toward these inci dents. Specifically, there is little research about whether correctional officers attribute bl ame to inmate victims of sexual assault and perceive their reports of rape as credible. In addition, research has failed to explore how correctional officers define rape in a correctional setting and thei r willingness to respond to these incidents. Victim Blaming Victims are often perceived as somewhat responsible for their own fates (Pugh, 1983), and attribution of blame to victims is an idea as old as victimology. Mendelsohn, considered the father of victimology, recognized that victims might be completely innocent, partially blamed, or fully blamed (Doerner & Lab, 2008). Research on attitudes toward rape victims began to emerge during the 1970s and has since been the subject of much empiri cal investigation. Although this topic, and rape research in general, has tr aditionally focused on fe male victims of male perpetrated crimes, research abou t male rape victims is beginning to emerge in social science research. The following discussion presents re search findings on percep tions of and attitudes toward female, male, and homosexual rape victims. Female Rape Victims Though it is important to understand which ch aracteristics may contribute to victim blaming, regardless of victim sex, most studies ha ve examined attributions of blame to female victims of male perpetrated rape. This rese arch suggests that pers onal characteristics and 30

PAGE 31

behavior of victims ma y influence belief in victim culpability. Some personal and relationship characteristics that have been found to increase vi ctim responsibility for rape in male perpetrated rape of females include revelati ons of past sexual conduct or rape victimizations and revelations of past or current intimacy be tween victim and defendant (Bell et al., 1994; Calhoun et al., 1976; Pugh, 1983; Schuller & Hastings, 2002). A few studi es also found that race could also influence blaming; Black victims and those in interraci al rapes are sometimes perceived as more responsible than others are (Foley et al. 1995; George & Martinez, 2002). Actions or behaviors of the victim that in fluence blaming include expressing emotion and behavior that contrasts with a genuine victim (such as not physically resisting and not reporting the rape immediately). Blaming is also a factor for victims perceived as provocative or victims who are believed capable of foreseei ng the incident (Calhoun, Cann, Selby & Magee, 1981; Kanekar & Kolsawalla, 1980; Krulewitz & Nash, 1980; McCaul et al., 1990; Spears & Spohn, 1996). In a review of judgments about female rape victims, Pollard (1992) concluded that male respondents were less likely to be pro-victim and that people who have less traditional attitudes about sex-roles were more pro-victim. One consistent finding is that males attribute more blame to female victims than female respondents do (Barnett, Quackenbush, Sinisi, Wegman, & Otney 1992; Calhoun et al., 1976; Cann, Calhoun, & Selby, 1979; Feild, 1978; Feldman-Summers, & Linder, 1976; Fulero & Delara, 1976; Kaneka r & Kolswalla, 1980; Kleinke & Meyer, 1990; MaCrae & Shepherd, 1989; Pollard, 1992; Thornton, Robbins, & Johnson 1981; Thornton & Ryckman, 1986; Thornton, Ryckman, & Robbins, 1982; Whatley, 2005). Pollard (1992) found that prior intimacy with an attacker reduces the perceived seriousness of the rape, and that carelessness by females increases blame attributed to female victims. Brownmiller (1975) and 31

PAGE 32

Burt (1980) suggest that attit udes toward female victims and victim blaming may extend from traditional patriarchal views about women as the minor sex. Research is somewhat mixed a bout how perceived respectabil ity of victim affects victim blaming. Jones & Aronson (1973) found that marri ed or virgin women victims were faulted more for their rapes than less respectable divorced female victims. Feldman-Summers & Lindner (1976), however, found that responsibilit y attributed to the victim increased and perceived impact of the crime decreased as respect ability (measured by marital status, or whether victim was a virgin or prostitute) decreased. In addition, male perpetrators are assigned less and female victims more culpability if the responde nts have high acceptance of rape myths, if resistance is offered late in the encounter, or if the defendant is una ttractive or has a less respectable occupation (Deitz & Byrnes 1981; Kopper, 1996; Smith et al., 1976). Traditional rape myths involving females as vi ctims are usually associated with attitudes toward female rape victims and beliefs about the culture of rape. Rape myths are derived from traditionally accepted gender roles and behavior. Rape myths about male perpetrated rapes of females include beliefs that female rape vic tims are lying, are usually promiscuous, or wear revealing clothing (Burt, 1980). Kopper (1996) found that respondents with low rape myth acceptance assigned less blame to victims and more blame to perpetrators. Jenkins and Dambrot (1987) found that those who agreed with rape myths assigned more blame to the victims, saw the victim as wanting sex, saw the perpetrator as less vi olent, and were not as li kely to describe rape scenarios as rape. Mason, Riger, and Foley (200 4) also found that a cceptance of rape myth increased victim blaming. Frese, Moya, and Megias (2004) found that level of rape myth acceptance interacted with situati onal characteristics to predict attr ibutions of victim blame and intensity of trauma: those with low rape myth acceptance attributed more blame and fewer 32

PAGE 33

traumas to victims who were acquainted with th eir assailants (compared to rapes involving strangers). Male Rape Victims Research about attitudes toward rape victims has traditionally focused on females as the typical rape victim, and although this depiction is accurate, studi es acknowledge that males are victims of rape as well. Research about males as victims of rape began to emerge in the 1980s, initially exploring topics such as female perpet rated rape of males and attributions of blame toward victims (Masters, 1986; Stuart & Gree r, 1984). Recent findings from the National Violence Against Women survey indicate that ab out 3% of men are victims of rape sometime during their lives, compared to the victimization rate of a bout 18% for females (Tjaden & Thoennes, 2006). One of the major problems in es timating the incidence of male rape is that only a minority of male victims report their assa ults (Davies & Rogers, 2006). Some research implies that male rape victims of male pe rpetrated rape are less likely to report their victimizations if they are acquainted with the perpetrator, however, most empirical evidence shows that all males are less likely than females to report their victimizat ions (Davies & Rogers, 2006; Hodge & Canter, 1998). The effects of male ra pe are often as serious and similar to that which females experience (King & Woollett, 1997; Perrott & Webber, 1996; Stermac et al., 2004), and include psychological eff ects such as depression, anger, loss of image, anxiety, and feelings of vulnerability, among othe rs (Walker, Archer, & Davies, 2005). Victim blaming in rape is often said to originate from varying sources, depending on the sex of the rape victim. While explanations of vi ctim blame with female rape victims are said to originate from feminist explanations that see th e world as patriarchal and hateful towards women (Brownmiller, 1975; Burt, 1980), blaming of male ra pe victims is thought to originate from sex role expectations (Dav ies & Rogers, 2006). Howard (1984a) f ound that male rape victims were 33

PAGE 34

more blameworthy when they acted outside of th eir expected sex roles (e .g., did not physically fight their attacker). Other research on male rape victims in the community suggests that the amount of resistance offered by males directly a ffects perceptions of blameworthiness (Howard, 1984b; Kassing & Prieto, 2003; Perrott & Webber, 1996). The traditional views of males as sexually aggressive and dominant, heterosexual and physically strong become the basis for male rape myths such as men cannot be raped (Davies, 2002). While many studies have examined acceptance of rape myths involving female rape victims, only a few studies have examined attitudes about male rape myths. Male rape myths are false beliefs and prejudicial stereotypes about the dynamics of male rape (Struckman-Johnson & Struckman-Johnson, 1992: 85-86). Research on th e culture of male rape indicates that males are expected to fight when threatened with rape and that people believ e that they cannot be forced to engage in sex without consent, cont ributing to the myth that males cannot be raped (Groth & Burgess, 1980). Two studies that sample college students found that women were less likely to accept myths about male rape (Chapleau, Oswald, & Russell, 2008; Struckman-Johnson & StruckmanJohnson, 1992). There is no research that explores inmate perceptions about male rape myths. Though the purpose of the current st udy is not to examine inmate beliefs about male rape myths, one purpose is to explore how correctional offi cers assume inmates feel about these issues. Victim blaming and attribution of credibility to male victims is similar to that of females (see above); however, in some instances male vi ctims are seen as less traumatized and more willing in acts of sexual assaul t and rape (Smith, Pine & Hawley, 1988). This effect is sometimes more pronounced among male resp ondents than among fe male respondents, 34

PAGE 35

suggesting that male rape victims may threaten traditional male ideals of masculinity and strength and expectations that males cannot be raped (Smith et al., 1988). Comparisons of Blaming for Fe male and Male Rape Victims Overall, research appears to be inconclusive about whether female or male victims are perceived as more blameworthy for their victim ization (Davies & Rogers, 2006; McCaul et al., 1990; Schneider, Soh-Chiew Ee, & Aronson, 1994). This research is also in conclusive regarding what types of statements about victims drive perceptions of blame. Howard (1984a, 1984b) found that males were more likely to be behaviorally blamed fo r their victimization (i.e., they were seen at fault due to their actual behavior, for example, failure to fight back), while females were more likely to be blamed for their characte r deficits (i.e., poor j udgment). She attributes this finding to the propensity to view male ra pe victims as acting outside the gender role expectations and attitudes at tributing traits such as physical strength to males. Anderson (1999), however, found that both male and female subjects attributed more behavioral blame to female victims rather than to male victims. Regarding characterological blame, male subjects attributed more blame to fe male victims and female subjects attributed this blame to both male and female victims equally (Anderson, 1999). Other research shows that male subjects were more likely than female subjects to attribute responsibility and pleasure to male rape victims (Mitchell et al., 1999; White and Robinson Kurpius, 2002). An interesting study by Janoff-Bulman, Timko and Carli (1985) found that subjects placed more blame on female victims in scenarios when the outcome wa s rape than when the outcome was neutral, and that this blame was char acterological in nature. In another study, heterosexual male subjects ma de the most anti-victim judgments of a male rape scenario, compared to heterosexua l female and gay male subjects (Davies & McCartney, 2003). In analyzing attitudes toward male and female rape victims, Burczyk & 35

PAGE 36

Standing (1989) found that males were more likely to attribute internal locus of causality (i.e., under the internal control of the victim) to victim s than to non-victims, and that female victims were viewed more sympathetically than male victims. Whatley & Riggio (1993) also found that male respondents attributed more blame to the victim than did female respondents. One study also revealed that gay male respondents were the least likely of the groups to endorse rape myths in the male rape scenario (i.e., the belief that a male could have fought o ff his attacker, or that most claims of rape are unfounded) (Davies & McCartney, 2003). It appears then, that ideas about blame may depend on factors other than victim characteristics, such as characteristics of research subjects as well as situational factors manipulated by the researcher. Homosexual Rape Victims Homosexual rape victims are viewed in some studies to be more at fault than heterosexual victims are (Davies & Rogers, 2006), es pecially when the perpetrator is a portrayed as a male (Davies et al., 2006). Mitchell et al. (1999) found that more responsibility and pleasure, and less trauma is attributed to homose xual victims of male rape than to heterosexual victims of male rape. Male rape victims who are homosexual are sometimes perceived to suffer less trauma than heterosexual victims are (D oherty & Anderson, 2004; M itchell et al., 1999). One study showed that in a hierarchy of trauma, heterosexual males were assumed to experience the most suffering, compared to heterose xual females and homosexual males (Doherty & Anderson, 2004). In another study, male subject s tended to be more pro-victim when the victim was portrayed as heterosexual as when the victim was portrayed as homosexual (Davies et al., 2006). Other studies, however, do not find that situat ions involving homose xual male victims are classified as rape as often as when the victim s are heterosexual women (Ford et al., 1998). In 36

PAGE 37

this same study, heterosexual females and homosexual males were seen as more responsible for their victimizations than homosexual females or heterosexual males (Ford et al., 1998). Burt and DeMello (2002) found that res pondents who scored higher on a homophobia scale attributed more blame overall to homosexual victims of rape, attributed more behavioral and characterological blame to heterosexual male victims, and attributed the least amount of blame to female victims. Anderson (2004) found that male subjects who had more homophobic attitudes had negative attitudes toward male rape victims. White and Robinson Kurpius (2002) also found that negative attitudes toward homosexuals related to more blame being attributed to homosexual victims. One study found that victims were attributed more blame if they were considered sexually attractive to the perpet rator, and gay men were attr ibuted more blame when the perpetrator was a male than were heterose xual males (Wakelin & Long, 2003). Also, blame toward gay males increased with greate r homophobic attitudes (Wakelin & Long, 2003). Regarding perpetrator blame, those victimizing ga y males were seen as th e least blameworthy of all perpetrators (Wakelin & Long, 2003). The character of homosexual men was a greater contributor to overall blame than the characte r of other rape victims (Wakelin & Long, 2003). Overall, research shows that se xual orientation is a major victim characteristic that affects blame attributed toward male rape victims. While most research analyzing perceptions of rape victims focuses on non-incarcerated victims, a handful of studies ha ve recently examined how correct ional officers view incarcerated victims of sexual assault and rape In the first study of its kind, Eigenberg (1989) found that forty-six percent of officers thought that inmates who had previously participated in homosexual acts in prison deserved to be raped (Eigenber g, 1989). In addition, 12-17% of those surveyed 37

PAGE 38

reported that inmates who acted like homosexuals or were homosexuals deserved to be raped, while 23% believed that previous involvement in homosexual activity caused an inmate to deserve rape (Eigenberg, 2000a). In the only st udy to examine influences of blaming in a correctional environment, officers who condemned homosexuality were more likely to blame victims of institutional sexua l assault (Eigenberg, 2000a). Victim Blaming by Inmates Though the current study does not explore inma te attitudes about rape victims, one purpose is to examine correctional officer percep tions of whether inmates blame other inmate victims of rape. There has been practically no research on inmate te ndencies to blame other inmates for their sexual victimization in correct ional institutions. One study of the social interaction of sexual assault in prison, however, suggest s that inmates percei ve those who violate the expectations of masculinity (e.g., by appear ing weak or vulnerable) as obvious targets for rape (Smith & Batiuk, 1989). Whether or not th ese inmates feel that this behavior merits victimization is unknown. Anot her study found that there is a tendency among inmates to assume that victims somehow deserve their fa te by participating in activities (like gambling and getting into debt) that make them vulnera ble to victimization (Jones & Schmid, 1989). In fact, Dumond & Dumond (2002) suggest that one of the reasons male inmates fail to report rape victimization is the fear that they will be blamed for the incident. In an analysis of interviews with inmates, Fleisher (2005) reports that inmates believe that all acts of sex in prison are willing acts, and that male inmates cannot be raped; however, inmates intervie wed recognized norms of behavior that were derived from prison rape lore. For example, inmates recognized the importance of avoiding debts, s uggesting that they may recognize possible scenarios for quid pro quo sexual victimization (Fleisher, 2005). 38

PAGE 39

Credibility of Rape Victims Contextual factors may influence perceptions of credibility or believability in reports of rape. Law enforcement officials consider legal variables such as physical evidence and witness accounts in allegations of sexual assault made by non-incarcerated victims, (Amir, 1971; Campbell & Johnson, 1997; Feldman-Summers & Palmer, 1980; Lafree, 1981; Weis & Borges, 1975). However, judgments of victims also influence whether their reports of rape are viewed as credible. How victims report the incident and the victim-offender relationship may also influence whether officers believe reports of rape (Estrich, 1987; Feldman-Summers & Palmer, 1980; Jordan, 2004; Weis & Borges, 1975). Some female victims are seen as less credible than others and some police investigat ors still view allegations of ra pe with suspicion and disbelief (Jordan, 2004). Female rape victims are seen as less credible if they are drunk or stoned, delay reporting, have psychological problems, or have previously consented to sex with their accused (Brownmiller, 1975; Jordan, 2004; LaFree, Resk in, & Visher, 1985; Schuller & Stewart, 2000; Torrey, 1991). In predicting what respondent characteristics influence perceptions of credibility, studies indicate that female subjects may be more prone than males to believe victim accounts of sexual assault (Bottoms & Goodman, 1994; Goodman, Bottoms, Herscovi ci, & Shaver, 1989; Jackson & Nuttall, 1994; ODonohue, Elliott, Nickerson, & Valentine, 1992). Other research indicates that occupational status may infl uence whether respondents attribute credibility to claims of rape; one study found that professionals, compared to students, were le ss likely to believe child victims of sexual assault (O Donohue, Smith, & Schewe, 1998). However, ODonohue, Smith, & Schewes (1998) research included elementa ry school teachers and university students; therefore, it is uncertain whet her the effects of occupation in their study may be partially explained by education level or age. 39

PAGE 40

Although official perceptions of the credibility of those re porting rape may be improved by available physical evidence, the purpose of this an alysis is not to determine the relevance of legal variables in determining the credibility of inmate s who report rape. One reason for this is that correctional officers to whom inmates report vict imization may not always be the staff members who will determine whether the incident is subs tantiated based on physical evidence or witness accounts. In other words, only a few officers in an institution ma y be charged with investigative functions to determine the nature of such alle gations and to recommend resolution. One purpose of the current study, rather, is to determine the re lative credibility of repo rts of rape by inmates with certain physical, character, or behavioral attributes. A common theme in research about sexual assault in prisons is that pr isoners rarely report their victimization (ODonne ll, 2004). This is not surprising, as a small percentage of males in the community report their victimizations to the police, and often, male cl aims of rape are not taken seriously by law enforcement officials (Davies & Rogers, 2006; King & Woollett, 1997; Mezey & King, 1989). Eigenbergs research (1989) uncovered some reasons why inmates may not report their experiences to staff. Officers repo rted that they are not likely to believe certain reports of rape, especially those coming from mu scular men, gang members, inmate leaders, and prostitutes. They were more apt to be lieve the young, those in debt, and sometimes homosexuals, but less prone to believe muscular inmates, gang members, or inmate leaders. About 31% either would never or rarely believe victims of ra pe who were prostitutes, 27% reported they were hesitant to believe homosexual rape victims, half were hesitant to believe muscular inmates, and two-thirds were hesita nt to believe inmate leaders (Eigenberg, 1989). Definitions of Rape In a 1991 survey of correctional officers in the Midwest, Eigenberg (2000a) found that correctional officers varied their de finitions of rape according to situational characteristics. They 40

PAGE 41

were less likely to define a s ituation as rape when inmates were coerced, were considered snitches, or were indebted to the perpetrator th an when inmates were physically threatened or overpowered. The literature on rape victims in th e community also suggests that criminal justice officials may use situational context (e.g., whether force was used and victim behavior) as well as legal cues (e.g., whether the victim consented) to define incidents as rape (Amir, 1971; Campbell & Johnson, 1997; Feldman-Summers & Palmer, 1980; Weis & Borges, 1975). Eigenberg (2000a) found that officers were more likely to define assaultive situations as rape if they avoided blaming victims of rape for their victimization. She also examined how officer attitudes about inmates and their profession influenced th eir definitions of rape. Her research indicates that officers who prefer social distance in their relationships with inmates and who are less concerned about th e ability of inmates to corrupt their authority were more restrictive in defining incidents as rape. Willingness to Respond Many officers believe that inmates do not re port incidences of rape to prison staff (Eigenberg, 1989). One of the reasons that inmates fail to report is that a reputation as a snitch can lead to further sexual or physical assault (Dumond, 2000; Fishman, 1934; Wooden & Parker, 1982). Another reason for inmate failure to report could be due to officer response to reports. Eigenberg (2000b) found that 8% of officers she surveyed reported ignoring inmate on inmate sexual behavior. Wooden and Parker (1982) also found that prison staff would ignore sexual activity unless it caused di sruption. One possibility for this is that officers may find it difficult to understand the differences between consensual, coerced, and forced acts of sexual activity. Almost all officers in one study repo rted that although they do not thi nk rape is rare, it is difficult to distinguish sexual acts as coercive or c onsensual (Eigenberg, 1989, 2000b; National Institute of Corrections & The Moss Group, Inc., 2006). 41

PAGE 42

Regarding responding to rape incidents, the few studies that have explored officer attitudes indicate that officers are willing to respond to prot ect inmates (Eigenberg, 1994, 2000b; Nacci & Kane, 1984). Eigenberg (1994) found that most officer s report willingness to respond to and prevent rape, though the types of responses they were willing to participate in varied. They reported moderate agreement with putti ng inmates in protective custody if they are pressured for sex, and about half reported having tolerance for inmates who fight when pressured for sex. In predicting willingness to respond to prison rape, those w ho embraced counseling aspects of correctional work, who pr efer less social distance in re lationships with inmates, have liberal attitudes toward women, condemn homose xuality, and are religious are more willing to respond. Officers with negative attitudes about victim s (i.e., that some deserve rape or that rape victims are weak) are le ss willing to respond to rape (Eigenberg, 1994). Officers may be less willing to use proactive responses for consensual homosexuality than for rape, and more apt to write disciplinary reports for sexual acts othe r than rape (Eigenber g, 2000b). In this study, however, officers showed greater support for patrolling areas a nd issuing disciplinary reports than they did for talking about the ri sks of sexual assault with inmates. Professional Orientation of Correctional Officers Professional orientation refers to overall outlook of correct ional officers about their job, their work environment, and inmates. A profe ssionally oriented officer is one who embraces job enrichment and the human service function of th eir job. They are not necessarily pro-custody, are willing to work to help inmates and improve prisons, are non-cynical, and are compassionate (Klofas & Toch, 1982; Toch & Klofas, 1982; Whitehead, Linquist, & Klofas, 1987). Counseling Roles, Rehabilitation, and Punitiveness Research on correctional officer attitudes towa rd the goal of prisons suggests duality in opinion about treatment and custody. Most rese arch indicates that while officers may be 42

PAGE 43

oriented toward custody as a goal of prisons and as a goal of their own jobs, they also recognize the importance of treatment and rehabilitation. However, in on e of the earliest studies of correctional officers, Duffee (1974) found that of ficers were most uncomfo rtable with social climates that show more concern for inmates (via rehabilitation and reinte gration) and were more comfortable with climates that focus on con cern for the community and prison management. However, around the same time, Jacobs & Kraf t (1978) found that guard s favor rehabilitation, punishment, and protection of society as the pr imary reasons for imprisonment. A recent study of both jail and prison staff suggests that correc tional officers in both types of institutions rank incapacitation over deterrence, reha bilitation, and retribu tion as the primary goal of corrections (Kifer, Hemmens, & Stohr, 2003). Officers worki ng in the prison were more supportive of the rehabilitative function than were officers working in the jail (Kifer et al., 2003). The majority of recent empirical evidence shows that correctional officers embrace treatment and rehabilitation as well as punitive goals of incarceration (Cullen, Lutze, Link, & Wolfe, 1989; Eigenberg, 2000a; Shamir & Drory, 1981). Specifically, studies show that many correctional officers disagree th at rehabilitation programs are a waste of money, and support rehabilitation programs that exist in prisons (Far kas, 1999; Toch & Klofas, 1982). It is important to note, however, that officers are more hesita nt to accept counseling as one of their own job roles (Farkas, 1999). An interesting finding of research on correctio nal officer attitudes suggests that officers believe their colleagues are much more procustody and opposed to work beyond custody roles than they actually are, a term re ferred to as pluralistic ignoran ce (Cullen et al ., 1989; Klofas & Toch, 1982). Pluralistic ignorance can be defined as the phenomenon in which members of a group systematically misperceive the attitude s and beliefs of their fellow group members 43

PAGE 44

(Kauffman, 1981: 273). Kauffman (1981) found that officers tend to view their colleagues as less sympathetic and less open to treatment of in mates than they actually are. Kauffman (1981) also found that officers perceive a consensu s among fellow officers that does not exist. Concern with Corruption of Authority Corruption of authority is a term first in troduced by Sykes (1958) and refers to the exchange relationship that exists between correc tional officers and inmates that is necessary in successful management of a prison. Research has attempted to operationalize corruption of authority by examining the extent to which o fficers trust inmates and their attitudes about professionally managing inmates (Whitehead et al., 1987). Eigenberg (2000a) and Toch & Klofas (1982) found that correctional officers express concern over the corruption of authority as well as a hesitancy to trust inmates. Toch and Klofas (1982) also found that many officers agreed that being lenient with inmates causes inma tes to take advantage of an officer, and most agreed that a good principle is to not get close to inmates. However, in this same study, many officers also disagreed that being firm and distant is the preferab le way to manage inmates (Toch & Klofas, 1982). Social Distance Correctional officers are favorable to keeping social distance between themselves and inmates (Eigenberg, 2000a; Farkas, 1999; Kasse baum, Ward, & Wilner, 1964; Shamir & Drory, 1981; Toch & Klofas, 1982). Empi rical evidence also suggests that they believe in acting respectfully and compassionately toward inmates (Farkas, 1999; Toch & Klofas, 1982). In a jail context, however, attitudes about the social distance between officers and inmates may be influenced by the career outlook of correctional officers. Pogr ebin and Poole (1988) found that jail correctional officers who used the positi on as a stepping-stone to a career as a law enforcement deputy were more interested in ke eping social distance between themselves and 44

PAGE 45

inmates than officers who placed more importance on the correctional officer job. The officers who perceived the correctional offi cer job as important and the jail as vital to the overall mission of the agency were more likely to indicate that knowing inmates was integral to successful job performance (Pogrebin & Poole, 1988). Empirical evidence suggests that correctiona l officers stress both treatment and custody functions of a prison, are concerned with corrupti on of authority, and contem plate social distance in their relationships with inmates. They show concern over the ability of inmates to corrupt their authority and their own abili ty to maintain sufficient social distance between themselves and inmates. Officers, however, may have some misperceptions about the punitive or custody orientations of th eir fellow officers. Jail Culture and Context One of the main reasons for conducting the curre nt study in a jail cont ext is because they are the most common correctional institution in th e United States. According to the most recent Census of Jails conducted by the Bureau of Justice Statistics, in 1999 there were 3,365 jails operating in the country (Stephan, 2001). In c ontrast, there were an estimated 1,668 state and federal prisons in 2000 and 1,821 prisons in 2005 (Stephan, 2008). At midyear 2007, jails in the United States held 780,581 inmates (Sabol, 2008a). This number, however, pales in comparison to annual jail admissions. Jails have the most significant percentage of turnover among inmate populations, and it is estimated that 13 million inmates were booked into jails in the United States during the 12 months preceding J une 2007 (OToole, 1997; Sabol, 2008a). Prison admissions during the year 2006, in contrast only totaled 749,798 inmates (Sabol, 2008b). Jails are often misunderst ood by the public, and many fail to understand the difference between a jail and a prison. Nearly 30 years ago, a sociologist identified the jail as a neglected area of research and stated everyone knows what a jail is until one begins to inquire into its 45

PAGE 46

definition (Mattick, 1974: 777). A jail is a locally operated correctional f acility that confines people before or after conviction (Schmalle ger & Smykla, 2007: 215). Inmates who are sentenced to jail usually have a sentence of less than one year (Clear et al., 2006; Smalleger & Smykla, 2007). The majority of jails are operate d under the jurisdiction of local sheriffs, though there are exceptions to this management st ructure (Mattick, 1974; Miller, 1978; Mays and Thompson, 1988). Five states assume the opera tion of local jail faci lities (Sabol, Minton, & Harrison 2007), and private companies operate 47 jails in the United States (Stephan, 2001). Most of the jails currently operating in the Un ited States are considered small jails, with capacities below 50 inmates (Kerle, 1998; Sabol, 2008a). Although jails operate under loca l administrations and vary m oderately in their facility design, resources, staffing, and proc edures, there are similarities across many jails in America. As previously noted most jails are operated by local sheriffs and compete with law enforcement as the primary mission of these department s (Mattick, 1974; Mays & Thompson, 1988; Miller, 1978; Pogrebin & Poole, 1988; Poole & Pogrebin 1988). Jail staff management, similar to prison staff management, is typically militaristic and hierarchical, with chain of command and personnel ranks. Jails are also si milar in that they serve multiple functions and therefore house a diverse population of inmates, from the disrupt ive drunk to the accused murderer (Klofas, 1990; Mattick, 1974). Functions of th e jail include: holding pre-trial detainees, probation and parole violators, juveniles, mentally il l persons, persons in contempt of court, court witn esses, prisonreleased inmates, inmates transferring to othe r facilities, and inmates due to overcrowding in other facilities (Clear et al ., 2006; Schmalleger & Smykla, 2007). Jail populations also share some similarities A large number of jail inmates have a mental health deficiency or are under the influe nce of drugs or alcohol at the time of admission 46

PAGE 47

(Abram & Teplin, 1991; Clear et al., 2006; G uy, Platt, Zwerling, & Bullock, 1985; OToole, 1997; Teplin, 1990, 1994). Facilities are typica lly crowded or overcrowded, suffer from relatively high suicide rates, and have insuffici ent fiscal resources, inadequate services, and inadequate programs for inmates (Bureau of Ju stice Statistics, 2008; Clear et al., 2006; Goss, Peterson, Smith, Kalb, & Brodey, 2002; Kimme, 1988; Klofas, 1990; Miller, 1978; OToole, 1997; Tartaro, 2003; Welsh, 1992). There are importa nt similarities across jails in the problems they face recruiting and training jail staff a nd offering adequate pay (Mattick, 1974; Miller, 1978; Schmalleger & Smykla, 2007). A prevailing pe rception of low job prestige characterizes the position of jail corre ctional officer and high staff turnove r and shortages exist in many jails (Kimme, 1988; Schmalleger & Smykla, 2007). There are also visible and important diffe rences between jail facilities, and these differences appear most visible when comparing small jails located in primarily rural areas to large jails located in primarily urban areas. A lthough facility size is indi cated by either inmate capacity or population count, and f acility setting is indicated by geographic location of the jail, the majority of small jails with inmate capacities of less than 50 inmates are located in rural areas (Ruddell & Mays, 2007). Although larger jails may be located in suburban as well as urban areas, this review dichotomizes jails by size and se tting and will refer to jails as small and rural or large and urban as much of the existing resear ch also dichotomizes jails in this manner. According to the Bureau of Justice Statistics, the number of small jails (i.e., jails with fewer than 50 inmates) decreased and number of large jails (i.e., jails with 1,000 or more inmates) increased during the period 1999-2005 (Sabol, 2008a). Overall, there is a larger percentage of ja ils considered small operating in the United States, however, jails with more than 1,000 inmates house slightly more than half of the nations 47

PAGE 48

jail inmates (Clear et al., 2006; Sabol, 2008a). Although there are 2,876 jail jurisdictions in the United States, 50 of the largest jail jurisdictions hold about a third of all jail inmates (Mattick, 1974; Sabol, 2008a). Though these jurisdictions are located across the country, (e.g., Los Angeles, CA; New York, NY, Cook County, IL, Harris County, TX, and Maricopa County, AZ), many are also located in Florida (Dad e County, Orange County, Hillsborough County, Jacksonville City, Pinellas County, Pa lm Beach County, and Polk County). Inmate crowding has been associated with bot h small and large jails, and recent research indicates that while crowding is still considered a problem for sm all rural jails (Ruddell & Mays, 2007), it may be more a problematic issue for jail s located in urban areas designed for greater inmate capacities (Applegate & Sitren, 2008; Be ck, 2002; Mattick, 1974). Sheriffs from large jails in one study were significan tly more likely than sheriffs in small jails to report that overcrowding influenced other problems in their facility (Kinkade, Leone, & Semond, 1995). Larger urban jail facilities also appear to be more likely than smaller jails to offer programs and services for the inmate population. St udies report that larger percentages of urban than rural jails have work, education, and trea tment programs available to inmates (Applegate & Sitren, 2008; Mays & Thompson, 1988; Ruddell & Mays, 2007). In addition, urban jails are more likely to have on-site or more sophisticated medical resources than rural jails (Applegate & Sitren, 2008; Kerle, 1982; Mays & Thompson, 1988; Ruddell & Mays, 2007). Other research studies indicate that smaller, rural jails have more difficulty providing services for inmates with mental health needs (Kellar, Jaris, & Manboah-Roxin, 2001; Mays & Thompson, 1988; Ruddell & Mays, 2007), although the need fo r these services in small, rura l jails may be significantly less than the need experienced by large, urba n jails (Powell, Holt, & Fondacaro, 1997). 48

PAGE 49

Accommodating juvenile inmates and housing them separately from adults may also be more problematic for small, rural jails (M ays & Thompson, 1988; Ruddell & Mays, 2007) The overall quality of the facility and availa ble resources differ according to jail size and setting as well. Small, rural jails are usually ol der and may therefore have greater deficiencies in their physical condition (Applegat e & Sitren, 2008; Kerle; 1 982; Klofas, 1990; Mattick, 1974; Mays & Thompson, 1988). Mattick (1974) argues, how ever, that jails that suffer from serious overcrowding, regardless of size, will age faster than the chronol ogical rate due to over-extended facilities. Small jail administrators in one study identified a lack of resources as problematic more often than large jail admi nistrators (Kellar et al., 2001) a nd have also reported a lack of fiscal resources as one of their most serious operational problems (Ruddell & Mays, 2007). There is inconsistent evidence regarding jail cont ext and legal problems; some research indicates that lawsuits are more problematic for small, ru ral jails and that they have more significant problems dealing with legal issues (Kimme, 1988; Miller, 1982), wh ile a more recent study reports that large, urban jails are more likel y to be under a court or der (Applegate & Sitren, 2008). A few studies report that sm all, rural jail facili ties are less likely to have problems such as inmate suicides (Kimme, 1988) and assaults on staff (Applegate & Sitren, 2008), however Mays and Thompsons (1988) research indicates that death, suicide, a nd homicide rates are higher in small jails than large facilities. The characteristics of staff members appear to vary across facility size and setting as well. Though small, rural jails have fewer sta ff members overall, resear ch suggests that they have more staff members per inmate and greate r proportions of female personnel than larger, urban jails (Applegate & Sitr en, 2008; Kellar et al., 2001). Personnel problems such as 49

PAGE 50

inadequate training and pay are more likely to be reported as significant pr oblems in small, rural jails (Kellar et al., 2001; Kerle, 1982; Miller, 1982). While jails of all types report similar problems and issues, it appears th at facility size and setting may either exacerbate or lessen the signi ficance of these problems. Smaller jails appear disadvantaged by fewer programs and services for inmates and older facilities in poorer condition. Smaller facilities are also challenged in accommodating the needs of special inmate populations such as juvenile inmates or those with mental health problems. Larger jails, however, appear to suffer from crowding or overcrowding more than smaller facilities (Applegate & Sitren, 2008; Beck, 2002; Mattick, 1974). Give n these effects of jail size, however, initial results from inmate self-reports in jails indicates that rates of sexual victimization do not differ signifi cantly by facility characteristics such as size or crowding (Beck & Harrison, 2008). Official documents and hearings with jail administrators, however, suggest that some facility characteristics (such as crowding, staff training, and staff turnover) might distinguish jails with high rates for sexual violence from jails with low rates. It is not yet known whether staff attitudes about se xual victimization differ across facilities that differ across characteristics such as size. Conclusion While many questions have yet to be explored in research examining jails, perceptions of jail correctional officers, and attitudes toward sexual assault victims, the literature reviewed sheds light on our understanding of these topics to date. While it appe ars that correctional officers in prisons are sophisticated in their unde rstanding of their diverse job functions and the competing goals of correctional institutions, they also have diverse perceptions about the nature of inmate-to-inmate sexual assault. This rese arch will explore the topics of sexual assault among inmates, perceptions of victims of sexual assault in correctional institutions, and the professional 50

PAGE 51

51 orientation of jail correctional officers in much more detail. The following chapter discusses the research project in more detail, specifically highlighting the methodology and the data used to explore these topics. Descrip tive statistics for the independe nt and dependent variables are presented in Chapter 4. Chapter 5 presents results of the analysis of perceptions of victims, specifically, attitudes toward vi ctim blaming and the credibility of inmates who report rape. Chapter 6 discusses results regarding how offi cers define rape in a jail context and their willingness to respond to consensual sex and sexu al assault among inmates. Chapter 7 presents the pluralistic ignorance findings to determine if o fficers accurately perceive the attitudes of their fellow officers and whether they believe they are mo re pro-victim than inmates in their facilities are. Finally, Chapter 8 summarizes the findi ngs of this research, its contributions and limitations, as well as suggestions for future research.

PAGE 52

CHAPTER 3 DATA AND METHODOLOGY Introduction I measured jail staff attitudes toward in mate-on-inmate sexual victimization and professional orientation by survey ing correctional officers in jails that house male inmates in Florida. The research design is cross-sec tional, and a non-probability purposive sampling method was used to capture the unique experiences of jail correctiona l officers. The nonprobability sample may not be re presentative of the population of correctional officers in Florida or in jails sampled, and this approach cannot gua rantee that the sample ch aracteristics will match the population characteristics (Kemper, Stri ngfield, & Teddlie, 1998). Although non-probability samples are derived from a population whose el ements have an unknown probability of being included in the sample, they are useful for seekin g information about specific research questions (Bachman & Schutt, 2007; Kemper, et al., 1998). The purpose of this research is not to generalize or draw conclusions about jail correctional officers in all settings. However, the study examines questions neglected by soci al scientists, so the purpose of the research is exploratory. In this case, the sample was not random. It wa s drawn from jails in Florida that agreed to participate in the research and granted access to their facility and staff members. Because the methodology was driven by the research question and setting, the sampling in this study is a type of purposive sampling method (Bachman & Schu tt, 2007; Maxfield & Babbie, 2001). The sampling strategy is also a type of availability or convenience sample, in that it selected research settings that were amenable to the researc h, were easily accessed, and sought to explore relatively unexamined research topics (Bach man & Schutt, 2007; Kemper et al., 1998). Completion of the project requ ired the cooperation of administrators in Floridas jails, and I contacted all county jail ag encies in the state of Florida. Although I initially secured the 52

PAGE 53

verbal approval from 15 of 67 county jail agencies, only 13 jail agencies participated in the research; a facility response rate of 19.4 percent. As discussed in more depth later, correctional officers at participating jails were given self-administered surveys to complete voluntarily and at their discretion. Some research has shown that self-administered questio nnaires are better than interviews at eliciting truthf ul responses to sensitive information (Acquilino, 1994; Jones & Forrest, 1992; Turner, Lessler & Devore, 1992); theref ore, I believe that this method is useful to obtain information on perceptions of sensitive issues such as attitudes toward sexual assault and sexual assault victims. Research Questions The first purpose of this research is to exam ine correctional officers attitudes regarding sexual assault and victims of sexual assault in ja il. The specific research questions include: 1. How do correctional officers perceive victims of sexual assault in jail? Specifically, do they attribute blame to victims of ra pe or sexual assault (victim-blame)? 2. How do officers perceive the credibility of vi ctims of sexual assault in jail? Are all inmates equally credible as victims of sexual a ssault, or are there ce rtain inmates that are more likely to be believed as victims? 3. How do officers define rape in a jail atmosphere? Specifica lly, what are their attitudes regarding sexual consent af ter coercion or threats? 4. How do officers perceive their own willingness to respond to acts of sexual assault and rape in jail? The second main purpose of the research is to examine correctional officer perceptions about the professional orientation of their fello w officers and beliefs about victim blaming and acceptance of rape myths among inmates. Th ese specific research questions include: 5. Are correctional officers in jails in Florid a accurate in their perceptions about the professional orientation of ot her officers? Specifically, how precisely do correctional officers identify their colleagues beliefs rega rding counseling roles, punitive orientation, corruption of authority, and soci al distance with inmates? 53

PAGE 54

6. How do correctional officer perceive the attitudes of inmates in their facility regarding victim blaming and belief in male rape myth s? How do these perceptions compare to their own attitudes of victim blaming and male rape myth? Settings Solicitation of Sites I began contacting jail administ rators in Floridas 67 county jail agencies in July of 2007 through email if the county jail agency or sheriffs office provi ded an email address on its web site. Otherwise, I made an initial telephone inquiry to a member of the jail administrative staff. I conducted follow-up contacts with a ll facilities that di d not initially respond to an email or telephone inquiry. All follow-up co ntacts were made via telephone; th erefore, if administrators were unresponsive to initial email inquiries I a ttempted to offer a verb al explanation of my research. During the initial contacts that were successful, I explained my graduate student status, the purpose of the research, and the methods I intende d to use to gather the data. At this time, I asked if the administrator would lik e to review the research material s (i.e., the survey instrument and consent form) in order to decide whether to participate. If an administrator agreed, I emailed, mailed, or faxed a copy of the consent form and the survey for further review. If these initial contacts were unsuccessful, I followed up until I had either: 1. spoken directly to a jail admini strator who was identified as a probable gatekeeper (usually this person was identified after explaini ng the study to a jail ope rator and/ or an administrative staff member), or 2. left two messages with a jail administra tor via an administrative staff member or voicemail. At the minimum, I contacted each of the 67 ja il agencies at least once via email and twice via telephone, or three times via telephone. By February 2008, 15 of the 67 jail agencies (22%) had expressed verbal agreement to participate. I continued contacts with the jails that either 54

PAGE 55

agreed to participate or were undecided through the end of the da ta collection period (September 2008). Facility Participation Only nine of the 67 agencies initially contacte d directly declined to participate (13%). Administrators in these nine ag encies expressed the decision not to participate mostly verbally over the telephone or through wr itten email correspondence; however, one administrator declined to participate through a mailed business letter. Some of these administrators offered information about reasons behind their decision not to participate. One facility suggested that the issues in the survey were not relevant to that facility. Others were vague about their decision, indicating they had operationa l issues or issues with staff partic ipating. Others simply stated that they will not approve the rese arch or that the jail would not be able to participate. While 15 jails originally verbally agreed to participate, four of these jails failed to respond to further inquiries about the researc h. During the data collection period (June 2008 through September 2008), two other counties that in itially agreed to participate declined and cancelled after research trips had already been scheduled1. Four jails that had not initially expressed interest in participati ng eventually agreed to participat e. Overall, 13 jails (19.4% of Florida jail agencies) participat ed in the research study by allowi ng me to visit the facility and speak to correctional officers during briefings or pre-shift meetings, ha nd out the surveys to officers outside of briefings, or by leaving consent letters and surveys to be distributed to the officers (Table 3-1). 1 One facility requested that I delete half of the survey questions in order for them to participate, which I declined to do to preserve the integrity of the study. The second facility cancelled on the day of data collection due to the publication of a commission report on alleged abuses in the jail. The jail administrator indicated that they wanted to cease participation to eliminate all distr actions from correctional officers duties. 55

PAGE 56

The 13 jails that participated varied in setting and geogr aphical location, ranging from small jails located in counties of less than 15,000 people to large ur ban jails in counties of more than one million people (Table 3-2). Facilities that participated also varied in geographical location in the state of Florida. At least one jail in each region of the state participated. Participating agencies include: Alachua County Sheriffs Office (Gainesville), Clay County Sheriffs Office (Green Cove Springs), Char lotte County Sheriffs Office (Punta Gorda), Escambia County Sheriffs Office (Pensacola), Jacksonville City2 Sheriffs Office, Jefferson County Sheriffs Office (Monticello), Leon Count y Sheriffs Office (Tallahassee), Palm Beach County Sheriffs Office (West Palm Beach), Pasco Sheriffs Office (New Port Richey), Pinellas County Sheriffs Office (Clearwate r), Sumter County Sheriffs Of fice (Bushnell), Taylor County Sheriffs Office (Perry), and Walton County Department of Co rrections (DeFuniak Springs). Survey Methodology For most of the study sites, I traveled to th e participating county jails and introduced the survey to jail correctional officer s during shift briefings or pre-sh ift meetings. I targeted each shift reporting to the jail within one 24-hour period. Survey packets were distributed to certified correctional officers of all ranks reporting to the jail facility for shift briefings during the one 24hour period. For example, if a jail facility ha d two twelve hour shifts that covered one 24-hour period (one shift working from 7:00 am-7:00 pm and one shift working from 7:00 pm-7:00 am), I visited both of these shift briefings in order to distribute surveys to the certified correctional officers working both shifts dur ing that 24-hour shift period. In addition to the counties that agreed to pa rticipate by allowing me to talk to officers during briefings, four jails agreed to participat e under other terms. On e county allowed me to 2 The city of Jacksonville, FL and Duval county, FL operate under a merged local government system. 56

PAGE 57

talk to officers and pass out the surveys as they entered the jail facility prior to, but not during briefings. Other facilities did not hold briefings; one of these f acilities allowed me to tour the facility to meet with and distri bute surveys to all officers on the shifts. I met with supervisors at two other facilities that did not hold briefings and left surveys and consent letters for their staff to distribute to officers. Administrators at two other f acilities that did not hold briefings agreed to notify their officers about the research and forward my contact information for participants to contact me directly. Currently, no contacts or surveys from these facilities have been received, and these jails are not included in the count of 13 participating f acilities. In addition, because I did not physically visit these coun ties, talk to officers employed there, or distribute surveys, I excluded these two counties from this analysis. One survey packet was distributed to each certified correctional officer, regardless of rank. The survey packets included a survey (see Appendix A), an informed consent letter, and a stamped and self-addressed return envelope. Af ter introducing myself during the shift briefings, I briefly described the research project and consent letter, id entified the purpose of the study, confidentiality and risk, and inst ructions for completing the survey. I advised the correctional officers that they should read the informed consent letter, and if they chose to participate to place the completed survey in a public mailbox. I al so advised the officer s both in the briefing discussion and in the informed consent letter that the survey was voluntary and anonymous and that they should not write their names on the survey. I chose to allow the participants to mail th eir completed surveys rather than conduct the surveys on-site in order to increase access to jails and to limit disruption to the facilities that participated. Many of the jails that participat ed conduct pre-shift briefings immediately before 57

PAGE 58

the incoming shift relieves the ou tgoing shift. The return-mail survey methodology was chosen to avoid requiring participating o fficers to arrive early for work in order to complete the survey. It was also impossible to allow officers to comple te the survey during or after briefings without disrupting facility operation or re quiring outgoing shift officers to st ay late for work because of the research. I also chose to allow correc tional officers to mail their completed surveys themselves rather than submit them to their supe rvisors to limit actual or implied perceptions of coerciveness by supervisors. This methodology also limited the opportunity for supervisors to open the surveys and read participant responses, thus guaranteeing anonymity. Research on surveys disseminated in this manner suggests that one can expect return rates of between 30-50% (Dillman, 2000; Dillman, Dolsen, & Machlis, 1995), and that this method is preferable to mail surveys due to increased social exchange. Not all correctional officers who worked at the participating jails were surveyed. I conducted the research during one da y at each participating site. I chose this data collection method due to limited resources, the nature of the re search settings, and expl oratory nature of the research. The date of data collection at each jail was scheduled in accordance with the schedules of the facility and administrators. This means that correctional officers who were not scheduled to work during the data collection day (i.e., they were not on rotation or duty) were not included in the sample population. In addition, correctio nal officers who may have not reported to work during the data collection day due to sickness, vacation, paid leav e, or other reasons were not included in the sample population. Jail Characteristics There is a wide range in the number of correct ional officers employed at each facility. The jail with the smallest number of employed correctional officers has only 13 employed officers. Four other jails (38% of the sa mple jails) also employ less than 100 officers and two jails (15%) 58

PAGE 59

employed between 100-200 officers (Table 3-1). There are several jails employing larger number of officers included in the study. One jail (8%) employs over 400 officers and four jails (31%) employs between 200-300 correctional o fficers. One facility (8%) employs 860 correctional officers at its facility, a number much greater than other participating jails. As previously indicated, the participating jails vary in their geographic location and setting. Only two of the fac ilities (15%) are located in coun ties with fewer than 25,000 people (Table 3-2). Four facilities (31%) are in counties with populations between 50,000 and 200,000. Three facilities (23%) are lo cated in counties that had 200,000 to 300,000 people. One jail county (8%) has a population of just over 450,000 and two jail county (15%) populations are between 825,000 and 925,000. The remaining jail c ounty has the highest population of about one and a quarter million people (Table 3-2). Surveyed jails are located across the state of Florida, including regions of sout h Florida, the Tampa bay area, nor theast Florida, north central Florida, the panhandle, and west Florida. Participating jails also differ in the number of inmates housed at the facility. The smallest jail houses fewer than 100 inmates3, while the largest jail house s 3,128 inmates (Table 3-2). Three jails (23%) house between 100-499 inmates, and three jails (23%) house between 500-999 inmates. Three of the jails (23%) house be tween 1,000 and 1,999 inmates, and one jail (8%) houses about 2,800 inmates. In addition, the jails differ in the type of housi ng and inmate supervision. They have either podular housing or a mixture of podular housing w ith dormitories, and/ or linear cell housing. Podular housing consists of individual cells surrounding a co mmon living area or dayroom (Clear et al., 2006). Six fac ilities (46%) have podul ar housing only while three (23%) have both 3 The inmate counts reported refl ect one daily count taken around the time of the facility visit. 59

PAGE 60

podular and dormitory housing both. One facility (8%) has linear cell and dormitory housing, and two facilities (15%) have al l three housing styles. The ma in types of supervision at participating jails are either dir ect, remote, or both. Direct supe rvision exists when correctional officers have direct interaction with inmates a nd are not separated from them by locked doors or secure control booths (Clear et al., 2006). Remote supervision ex ists where correctional officers either observe inmates from a control room that is secured from inmate access or through locked cell doors (Schmalleger, & Smykla, 2007). None of the facilities indicated that they only intermittently supervise inmates. Five (38%) ja ils have direct supervision only, and five have remote supervision only. Two jails ( 15%) have both types of supervision. Sample A total of 1,161 surveys were distributed to officers at the 13 participating jails. According to jail administrators this represents 41% of the 2,822 total officers employed at these jails (or the target population) (Table 3-1). The sampled population of 1,161 officers differs from the target population of 2,822 because not a ll officers employed at the jails were working the shifts I visited. The total number of surveys received from officers in all jails surveyed is 376, a response rate of 32.4%4. The number of surveys received represents 13% of the target population of 2,822 officers. The number of surveys received from officers at these jails varied somewhat, as jails employing the most officers did not necessarily have higher response rates for the survey. For example, one of the jails that I surveyed that employs a small number of officers and houses fewer inmates has the highest response rate. Another small jail and a medium-sized jail have the second and third highe st response rates (Table 3-1). As expected, four of the largest 4 The response rate was calculated by dividing the number of surveys returned (376) by the number of surveys distributed (1,161). The response rate was on the low end of the expected 30-50% respond rate suggested by Dillman (2000). This could be a result of the sensitiv e nature of the topic and questions in the survey. 60

PAGE 61

jails surveyed have the largest numbers of surv eys distributed and the largest percentages of respondents in the total sample (45%, 8%, 11%, and 10%, respectiv ely) (Table 3-1). Subjects were asked to respond to several de mographic questions, including: sex (female = 1, male = 0), race (non-White = 1, White = 0), ethnicity (Hispanic = 1, non-Hispanic = 0), marital status (not married = 1, married = 0), education level, age, political party (nonRepublican = 1, Republican = 0), po litical orientation, religiosity, and whether a family member or friend had ever been incarcerated. More ma le correctional officers (69%, n = 260) responded to the survey than female correctional officer s (31%, n = 115) (Table 3-3). However, the proportions of male and female respondents matc hed those employed by these facilities. The majority of respondents are White (78%, n = 293) and non-Hispanic (95%, n = 356). Sixteen percent (n = 61) identified themselves as Black a nd less than 3% (n = 8) indicated membership to other racial categories. The proportion of Whites is larger an d the proportion of Blacks is smaller in the sample than in the target populatio n of officers employed at these facilities (Table 3-3). Most of the correctional o fficers in the sample have completed schooling beyond a high school education. Similar to th e target population of all officers employed by these 13 jails, no respondent indicated that they have less than a high school diploma or GED5. Almost half (48%, n = 181) do not have a college degree yet have received some college credit. Many reported obtaining either less than a year of college cred it (21%, n = 79) or more than a year of college credit (27%, n = 102) (Table 3-3). About 21% (n = 78) have completed an associates degree, while 19% (n = 71) have obt ained a bachelors degree. 5 The requirements of employment at thes e jails dictate that correctional officer s must have at least a minimum of a high school diploma or GED. 61

PAGE 62

In order to compare the educational levels of the sample population with the target population, I obtained data, where available, fr om the participating facilities about the educational background of their co rrectional officers. Six facilities employing 1,707 correctional officers (60% of the target population of 2,822) do not collect data on college degrees or college credit received by their employees. In the facili ties that do collect this information, only 27% (n = 304) of the total population of 1,115 correctional officers employed in these facilities have an associates degree and only 14% (n = 160) have a bachelors degree. It appears that the educational level of the sample is similar to that in participating facilities that collect data on educational level of their officers as about 21% of the sample reported having an associates degree and 19% of the sample repor ted having a bachelors degree. Less than 2% (n = 6) of the sample reported having a masters degree and less th an 1% (n = 7) report ed having a professional or doctorate degree. Likewi se, less than 1% of the population of 1,115 officers for which educational information was available (n = 7) have a Masters degree (Table 3-3). For the purposes of the later analyses, I recoded education into a dummy variable that distinguishes officers who have received a college degree (1) from officers who have not received a college degree (0). The majority of respondents are married (64%, n = 241), however, many are also divorced (17%, n = 62) or have never been married (1 6%, n = 60) (Table 3-3). The majority of respondents are between 31 and 45 y ears of age (56%, n = 208). Si xteen percent (n = 58) are in the 31-35 age category, 21% (n = 79 ) are in the 36-40 age category, and 19% (n = 71) are in the 41-45 age category. Only 11% (n = 44) replied that they are between the ages of 46 and fifty. The respondents are relatively split on political party and orientation, though slightly more respondents affiliate with the Republican Party or conservative political ideology. While 35% (n 62

PAGE 63

= 133) report that they belong to the Democratic Party, 41% (n = 155) re ported membership in the Republican Party. Many (17%, n = 63) are also Independents or affiliated with another party (5%, n = 17) (Table 3-3). Not many correctiona l officers reported bei ng liberal (7%, n = 26), very liberal (2%, n = 7), or ex tremely liberal (1%, n = 4). A large portion of respondents (43%, n = 163) indicated that they consider themselves middle of the road with regard to political orientation. Many also indicated that they are conservative (32% n = 119), very conservative (7%, n = 28), or extremely conserva tive (3%, n = 11). In addition, the mean of this variable is 4.44 (where one indicates extremely liberal and seven indicates extr emely conservative), reflecting a slight tendency for this sample to describe themselves as conservative over liberal (Table 3-3). Officers were also asked to indicate on a scal e of one (not at all) to 10 (extremely) how religious they considered themselves. Th e mean response was 5.76, and the response was moderately bi-modal with 76 respondents (20%) indi cating a score of eight and 74 respondents (20%) indicating a score of five (T able 3-3). These results indicate that the sample is somewhat to moderately religious. Forty-four percent (n = 163) indicated a score of seven or above on the religiosity scale. Only 26% (n = 96) replied to this question with a score of four or less. Respondents were asked whether any of their family or friends had ever been incarcerated, and the majority indicated yes to this question (67%, n = 253). The survey also requested information about job characteristics such as shift they typically work, job satisfaction, stress experienced on the job, years employed as an officer, and whether their facility is located in a rural, urban, or suburban area. In order to measure job satisfaction, officers were asked to indicate on a scale of one (not at all) to 10 (extremely) how satisfied they are with their job. Overa ll, the officers in the sample rate th emselves moderately satisfied with 63

PAGE 64

their job ( x = 6.92; mode = 8.00) and 65% (n = 243) ra te themselves with a seven or above (Table 3-3). Only 12% (n=45) rate themselves with a four or below on this scale. Officers reported moderate levels of stress ( x = 5.96; mode = 5.00) when asked to indicate on a scale of one (none) to 10 (extreme) the am ount of stress they experience on the job (Table 3-3). Only 23% (n = 85) indicated that their le vel of stress is a four or below. Twenty-one percent (n = 79) wrote in a five, 13% (n = 50) wrot e in a six, and 39% (n = 144) indi cated stress levels of seven or above on this scale. Although the survey asked officers to report wh at shift they usually work, there were problems with the answer options discovered af ter data collection ha d begun. The answer options included: day shift (morni ng to afternoon), evening shift (afternoon to late evening or midnight), and night shift (late evening or midnight to early morning). These answer options were originally based on personal correctional experience and were verified by an administrator in one of the participating jails prior to data collection. After data collection had begun, some of the correctional administrators advised me that thei r shifts were not organized as described in the survey. Rather, many facilities had 12-hour sh ifts that generally began between 5:00 am and 7:00 am and between 5:00 am and 7:00 pm. Some facilities had officers that rotate on and off shifts with varying work hours. I made the decisi on not to adjust the answ er options according to discussions with administrators in order to facilitate the completi on of the project and to maintain consistency across distributed and completed surveys. Regardless of this problem, only 18 respondents (5%) left this question unanswered. Of those who answered this question, 45% (n = 168) re plied that they typically work the day shift. Thirty-two percent (n = 120) indicated that they work the night shift, and 19% (n = 70) work the evening shift. Some officers (n = 10) circled two responses, whic h may suggest that they work 64

PAGE 65

on a 12-hour shift that they believ e includes two answer responses or a rotating shift schedule. Of those who circled two answer options, 80% circled day and night shift and 20% circled day and evening shift. Eleven of the jails returned information about th e number of officers assigned to each shift. Two jails6 indicated that they would not give out this information fo r security purposes. Of the eleven jails that indicated the number of officers per shift, two indicated that the number of officers on each shift is equal. One county has a shift rotation schedule that uses all of its correctional officers on rotating days and nights. Seven facilities have fewer officers assigned to the night through morning shifts than on the morni ng through night shifts. Of the counties that have two 12-hour shifts during a 24-hour period, most have fewer overall officers assigned to the night shift than the day shift. Two facilities have three 8-hour shifts during a 24-hour period. In one of these facilities, more officers are assigned to the 8-hour shift covering late afternoon th rough about midnight (211pm) than the other two shifts. The other facility has a larger num ber of officers assigned to the day shifts than the nigh t shifts. A third facility has a la rge number of both 12-hour and 8-hour shift assignments, and both shift assignments have more officers assigned to the day shifts than the night shifts. The information from the surv ey and provided by the facilities regarding the number of officers assigned to sh ifts suggests there are a slightly larger nu mber of correctional officers working during day sh ifts than night shifts. In order to determine length of experience, o fficers were asked to estimate to the nearest year how long they have been working as a co rrectional officer at any facility. The modal response is 2 years and the average response is 10 years. A similar number of respondents 6 Alachua County Sheriffs Office (employing 227 offi cers) and Taylor County Sheriffs Office (employing 33 officers). 65

PAGE 66

reported working as a correctional o fficer for either four years or less (26%, n = 95) or from five years to less than 10 years (26%, n = 94). Thir ty-six percent (n = 132) reported working as an officer for more than 10 but less than 20 years, and only 12% (n = 45) re ported employment as an officer for more than 20 years. Most of the respondents indicated that their facility is located in an ur ban area (70%, n = 241). Urban area was defined in the survey accord ing to the United States Census Bureau (2000) definition as an area with a to tal population of 50,000 or more. Rural area was defined in the survey as the open country or in a town of less than 2,500 people (Rockefeller, 1972; United States Census Bureau, 1995, 2000; United States Department of Agriculture, 2004). Only 9% (n = 31) responded that their jail facility is located in a rural area. For the purposes of this survey, suburban area was defined in the survey as s omewhere between rural and urban and about 21% of respondents (n = 71) indicated that thei r facility is located in a suburban area. Data Officer Perceptions of Victims Independent variables Among demographic and job char acteristics, the survey asked correctional officers questions in the following areas: professional orientation, attitudes towa rd homosexuality, and acceptance of male rape myths. Most of the questions asked the respondents to indicate their level of agreement on a 6-point Likert scale, with answer options including strongly agree, agree, not sure but probably agree, not sure but probably disagree, disagree, or strongly disagree. The professional orientation items originate from Klofas & Toch (1982), Toch and Klofas (1982), and Whitehead et al. (1987). Whitehead et al. (1987) found that the factor structure of items in the professional orientat ion items originally proposed by John Klofas and Hans Toch (Klofas & Toch, 1982; Toch & Klof as, 1982) adhere to four sub-topics when 66

PAGE 67

subjected to factor analysis. These questions, taken directly from Whitehead et al. (1987), asked respondents several questions a bout whether correctional offi cers should be involved in counseling roles. These statements included the following items: reh abilitation programs should be left to the mental health professi onals, counseling is a job for counselors, not correctional officers, and if a correctional officer wants to do counseling, he should change jobs (Table 3-4). Confirmatory factor analysis using varimax rotation was used to examine whether statements about counseling roles tapped one or more concepts. The factor loadings, displayed in Table 3-4, indicate that the statements are unidimensional, consistent with Whitehead et al. (1987). I created a counseling roles inde x by combining and averaging responses to these three items and found the index reliable ( =.77). Each index variable in this analysis was created by adding all response values across all variables in the index and dividing by the number of variables in the index for an average response. Res ponses to the counseling roles index are coded so that higher values reflect more suppor t for counseling roles and higher levels of professional or ientation (Table 3-4). The second professional orientati on sub-topic explored in this analysis is acceptance of punitive orientation. Officers were asked the foll owing three questions to gauge their level of punitive orientation: rehabilitation programs are a waste of time and money, improving jails for inmates makes them worse for officers, and a military regime is the best way of running a jail (Table 3-4). These statements were comb ined to form a punitive orientation index after confirmatory factor analysis with varimax rotation indicated a singl e concept. Although Whitehead et al. (1987) include a fourth item on their index measuring punitive orientation, the data in this analysis do not s upport reliability or c onsistency in measurement for a four-item index (Bachman and Schutt, 2007). The reliability for punitive orientation strengthens somewhat 67

PAGE 68

(from = .55 to = .58) by dropping this fourth item (th ere would be less crime if jails were more comfortable) from the analysis. Although f actor loadings of items on this index indicate valid measurement of punitive orientation, reliability is relatively weak. However, the index is used throughout the analysis because it has few items it is only used as an independent variable, and has been shown reliable in previous studies (Eigenberg, 2000a; Whitehead et al., 1987). Items on this index are coded so that higher scor es reflect greater punitiv e orientation and lower levels of professional orientation. The respondents were asked five questions that addressed th eir concern about the ability of inmates to corrupt their au thority. These statements incl uded the following items: you cant ever completely trust an inmate, a good principl e is not to get close to inmates, a personal relationship with an inmate inv ites corruption, you must keep c onversations with inmates short and business like, and if an officer is lenient with inmates they will take advantage of him/ her. Confirmatory factor analysis using vari max rotation determined that the items tapped one overall concept and the items were summed and averaged into a corrup tion of authority index (Table 3-4). This finding is similar to Whiteh ead et al. (1987) as the index is also reliable ( = .74). The items in the index are reverse-coded so that higher scores reflect more concern over corruption of authority or le ss professional orientation. The last professional orientati on sub-topic examined in this study is officer preference for social distance in their relations hips with inmates. Officers were asked the following five questions to measure preference fo r social distance: a correctional officer should work hard to earn trust from inmates, its important for a co rrectional officer to have compassion, you get to like the inmates you supervise, sometimes a correctional officer should be an advocate for an inmate, and the way to get respect from inma tes is to take an interest in them. These 68

PAGE 69

statements were combined to form a social di stance index after confirmatory factor analysis using varimax rotation indicated unidimensionalit y, consistent with Whitehead et al., (1987). The social distance index is moderately reliable ( = .67). The items in this index are reversecoded so that higher scores reflect more pref erence for social distance from inmates or less professional orientation. Herek (1984; 1988) developed what was orig inally called the C ondemnation-Tolerance index of attitudes toward homo sexuality, which was a combination of items from other studies (Levitt & Klassen, 1974; MacDonald, Huggins, Young, & Swanson, 1973; Millham, San Miguel, & Kellogg, 1976; Smith, 1971) as well as additional items Herek constructed. He argued that the index is unidimensional and more ge neralizable than previous indices used as it was tested on samples in several geographic areas. Items from Herek (1984, 1988) include questions such as male homose xual couples should be allowed to adopt children the same as heterosexual couples, I think male homosexuals are disgusting, and male homosexuality is merely a different kind of lifestyle that shoul d not be condemned (Table 3-4). The ten items used in this study were adopted from Hereks (1 988) Attitudes Toward Gay Men index, a shorter version of his (1984) Condemnation-Tolerance in dex. Eigenberg (1994, 2000a) used a modified version of Hereks Condemnation-Tolerance (1 984) index and found it both unidimensional and reliable. The ten items adopted for this study to test attitudes toward homosexuals loaded on one factor in a confirmatory factor analysis using varimax rotatio n. These items were summed and averaged to form a ten item index of attitudes toward homos exuality that was internally consistent ( = .94) (Table 3-4). Four of the items in this index are reversecoded so that higher scores on the index reflect greater acceptanc e of homosexuals and homosexual lifestyles. 69

PAGE 70

Male rape myth acceptance was measured using four separate items adopted directly from Struckman-Johnson and Struckman-Johnson (1992). These items include the following statements: most men who are raped by a man are very upset by the incide nt, it is impossible for a man to rape a man, most men who are raped by a man do not need counseling after the incident, and even a big, strong man can be raped by another man. These items are not reliable or unidimensional when subjected to reli ability or exploratory fa ctor analysis, therefore they are kept separate for the analysis (Table 3-4). Items are coded so that higher responses indicate acceptance of or adherence to male rape myths (i.e. the belief that it is impossible to rape males, etc.). Dependent variables There are four main dependent variables in the survey that corre spond to the research questions outlined in Chapter 3: victim blaming, credibility of victims, definitions of sexual assault, and willingness to respond to sexual assault incidents. The majority of the questions ask the respondents to indicate thei r level of agreement on the 6-poin t Likert scale. The following section will first discuss the two dependent variab les dealing with perceptions of inmate victims of sexual assault, specifically victim blaming a nd perceptions of inmate credibility. Following this, I will discuss the remaining two dependent vari ables: officer definitions of rape and officer willingness to respond to sexual assault in jail. The five questions regarding the extent of blame or responsibili ty that correctional officers attribute to victims of sexual assa ult were adopted from H. Eigenberg (1989, 1994, 2000b, personal communication, June 7, 2006), who reported a unidimensional and reliable factor structure. These five items include inmates who have previously consented to participate in sexual acts, homosexual inmates, inmates w ho take money or cigarettes in exchange for consensual sexual acts, and inmat es who dress or talk in femini ne ways get what they deserve 70

PAGE 71

if they are raped. The last item adopted from H. Eigenberg (1989; 1994; 2000b, personal communication, June 7, 2006) ask whether some inma tes deserve to be raped in jail because of the way they act. Two additional items on the victim blaming index were included from StruckmanJohnson and Struckman-Johnson (1992). These it ems include: most men who are raped by a man are somewhat to blame for not being more careful and are somewhat to blame for not escaping or fighting off the man. A reliability and exploratory factor analysis using varimax rotation of all seven items indicates that these items are both reliable ( = .93) and load on one factor (Table 3-5). A victim-blaming index was created by adding resp onse values across the items and dividing by seven for an average response Items in this index were reverse-coded so that higher scores reflect great er victim blaming attitudes. In order to measure the percep tions of credibility of victim s, officers were asked whether they agree that factors such as drug use and delay in reporting affect the credibility or believability of an inmate who reports that he was raped. They were al so asked the extent to which they believed victims with varying physic al attributes (e.g., youth or muscularity), and answer choices range from always believe to n ever believe. These items were adopted from Jordan (2004) and Eigenberg (1989). These item s were submitted to an exploratory factor analysis using varimax rotation and comb ined to form a highly reliable index ( = .91) that loaded on one factor (Table 3-5). Items were reverse-coded so that higher scores on the credibility of victim index reflect greater beli ef in credibility of inmates who report rape. Measures of definitions of rape include five short vignettes that portray sex between two inmates under varying circumstances. Respondent s indicated their leve l of agreement with statements like inmate Jones tells inmate Sm ith he will kill him unless Smith has sex with 71

PAGE 72

Jones, Smith has sex with Jones, Smith has been raped7. These items are also measured with a 6-point Likert scale that indicate s level of agreement or disagreem ent. Eigenberg (2000a) reports that these definitions represent one factor and are reliable. Th e results of the reliability and confirmatory factor analysis using varimax rotati on are consistent with th is prior study as the items are reliable ( = .87) and unidimensional (Table 3-5). The responses to these five vignettes were summed and averaged to create a rape defi nition index. Items in th is index were reversecoded so that higher scores reflect a te ndency to define the incidents as rape. Officers were also asked to report on thei r own willingness to respond by indicating how they believe officers should respond to incidents of consensual sex or sexual assault. Questions were derived from H. Eigenberg (1994; pe rsonal communication, June 7, 2006) and include items such as: officers should encourage inmates to report sexual assaults, officers should talk to inmates about the risk of sexual assault, and officers shoul d refer inmates to protective custody to safeguard them from sexual assault (T able 3-5). Eigenberg ( 1994) reports reliability and unidimensionality for this index; however, the data in this analysis do not support unidimensionality when items were submitted to confirmatory factor analysis. Though conceptually I assumed that items dealing with consensual sex would load together on one factor, while items related to sexua l assault would load as a separate factor, the data in this study also did not follow this pattern. For this samp le, items dealing with talking to inmates about these incidents load separately and rather strongly on one factor, indicating that officers view talking about these issues conceptually different from responding in other ways. The willingness to respond items load on thre e separate factors when submitted to an exploratory factor analys is using varimax rotation. The first f actor contains four items indicating 7 Though the wording of the questions was replicated exactly from prior re search, it is important to note that including the phrase Smith has been raped in each statement possibly biased the subjects responses. 72

PAGE 73

that officers should encourage inmates to report sex between inmates and that officers should do everything they can to prevent these incidents, both consensual and assaultive. Reliability analysis for these items indicates consistency ( = .70). These four items were reverse-coded, summed, and averaged to form an "encourage in mate reporting and officer prevention" index. Higher scores reflect agreement that officers shoul d adopt these responses to consensual sex or sexual assault among inmates. The second factor includes two items that s uggest that officers should talk to inmates about consensual sex and about the risks of sexual assault. These two items were reverse-coded, summed, and averaged to form a talk to inmates index that is highly reliable ( = .81). The final factor includes the two items that encourage proa ctive actions such as us ing cell assignments and protective custody to safeguard inmates from sexu al assault (Table 3-5). Although this item is unidimensional, it has a moderate measure of internal consistency ( = .67). These items were also reverse-coded and combined to form a proactive action index. Th e items on these indices are coded so that higher scores reflect agreement that officers should adopt that response to consensual sex or sexual assault among inmates. Perceptions of Fellow Officers Professional Orientation As previously noted, another goal of the re search is to examin e whether correctional officers in Florida jails accurately perceive the at titudes of their fellow officers. In addition to questions about their own profe ssional orientations, officers were asked to guess how other officers in their facilitie s would answer questions about profes sional orientation. For each item in the counseling roles, punitive or ientation, corruption of authority, and social distance indices, officers were asked to indicate whether almost al l (over 80%) or most (more than half but fewer than 80%) would agree or whether almost all (over 80%) or most (more than half but fewer than 80%) would disagree. 73

PAGE 74

Perceptions of Inmate Attitudes Finally, officers were asked to guess how inmates in their facility would respond to the questions about victim blaming and male rape myths previously discu ssed. For each victim blaming and male rape myth item, officers were asked to indicat e whether almost all (over 80%) or most (more than half but fewer than 80%) of inmates would agree or whether almost all (80% or more), or most (more than half but fewer than 80%) of inmates would disagree. Analysis Officer Perceptions of Victims The analysis of perceptions of victims of sexual assault in cludes examination of four dependent variables: victim blaming, credibility of victims, definitions of rape, and willingness to respond to consensual sex or se xual assault among inmates. The first part of the analysis for each of these dependent variables consists of a paired sample t -test to determine whether there are significant differences across the sample means for various questions in the blaming, credibility, definitions of rape, and willingness to respond indices8. Specifically, each item in each separate index is compared to each other it em in the same index to determine if there are significant sample differences in response to th e items. For example, each victim blaming statement is compared to each ot her victim blaming statement to determine if the sample means for blaming are significantly greater for one type of victim than for others. Paired t -tests are used because each officer in the sample was asked to indicate their response for each item; therefore, their responses for one item may be dependent on their responses for other items (Agresti & Finlay, 1997). 8 An alpha level of .05 is used for all statistical tests and all t -tests are two-tailed. 74

PAGE 75

The second part of the analysis for each de pendent variable meas uring perceptions of victims consists of bivariate correlations with the independent variables attitudes toward homosexuality, male rape myth acceptance, and th e professional orientat ion items of counseling roles, punitive orientation, corrupt ion of authority, and social di stance. Demographic and work variables are also initially incl uded in bivariate correlations for all four dependent variables. Specifically, these correlations in corporate age, race, sex, and e ducation as they have been shown to be important predictors of officer attitudes toward inmates in prior studies (Arthur, 1994; Crouch & Alpert, 1982; Cullen et al., 198 9; Farkas, 1999; Jackson & Ammen, 1996; Jacobs & Kraft, 1978; Jurik, 1985; Maahs & Pratt, 2001; Poole & Regoli, 1980a; Teske & Williamson, 1979; Toch & Klofas, 1982). The work variables job satisfaction, stress, and seniority are also included in the bivariate correlat ions as some prior research indicates that they are important factors in officer attitudes (A rthur, 1994; Cullen, Latessa, Burton, & Lombardo, 1993; Farkas, 1999; Jurik, 1985; Poole & Regoli, 1980a; Teske & Williamson, 1979). In some instances, the dependent variable s are also entered into bivari ate correlations with other dependent variables to determine whether they ar e significantly associated. Because perceptions of non-incarcerated rape victims influence blaming tendencies and believability (FeldmanSummers & Linder, 1976; Jordan, 2004; Kanekar & Kolsawalla, 1980; Kr ulewitz & Nash, 1980; McCaul et al., 1990; Pollard, 1992; Pugh, 1983; Spears & Spohn, 1996), and because several of the items in the dependent variable indices describe different types of victim s, it is expected that the dependent variables may affect each other. The variables that are significantly correla ted with each dependent variable are then entered into a multivariate regression analysis with that dependent variable. The multivariate analysis consists of theory-driven, forced-entry stepwise ordinary least squares (OLS) regression 75

PAGE 76

models that use demographics, work variables, and independent variable s to predict attitudes about victim blaming, credibility of victims, definitions of rape and willingness to respond. The rationale for using linear regression stems from the nature of the ordinal measurement of the dependent variables. I have check ed that the data are consistent with the major assumptions of linear regression: linearity, homoscedasticity9, normally distributed erro r terms, and lack of multicollinearity. The stepwise models are sp ecified using theoretical model building and a forced-entry method rather than computer generated model building. Stepwise linear regression allows analysis of R2 and coefficient change across equations to understand specific contributions and explanatory power of variab les. The baseline stepwise equation for each of the four dependent variable s include as independent variables demographic and work variables that were bi variately correlated with them. These variables are entered first to assess the contribution of the other independ ent variables after cont rolling for demographic and work variables. Each subsequent block inco rporates additional indepe ndent variables in the following order: attitudes toward homosexuality, professional orientation items, and male rape myth acceptance. Where applicable, the dependent variables that are significantly correlated with other dependent variables are included in the models. I chose the order of variable entry into the model to control for f actors that prior literature sugge sts are important while examining unique contributions of professi onal orientation items, male ra pe myths, and perceptions of victims not tested in prior studies. The pa rameters of each regression model (including 9 Examination of the scatterplot of the predicted values and residuals in all four dependent variable analyses indicates possible heteroscedasticity. Violation of the ho moscedasticity assumption results in deflated standard errors and the type I error of inapprop riately determining significance. Models are estimated with robust standard errors to correct for inflated standard errors. The results are identical to thos e without robust standard errors, with the exception of the definitions of rape analysis and one willingness to respond analysis (Studenmund, 2001). For these analyses, the results are presen ted using robust standard errors. 76

PAGE 77

coefficients, standardized coe fficients, standard errors, and t -values) are presented in table format in Chapters 5 and 6. In surveys, it is common to have data mi ssing from cases due to non-response (Allison, 2002; Dempster & Rubin, 1983; Littl e, 1983). Missing data can eith er be classified as missing completely at random (i.e., probability of data missing on one variable is unrelated to both its value and the value of any other variable), or missing at random (i.e., probability of data missing on one variable is unrelated to its value after controlling for ot her variables in the analysis) (Allison, 2002). Cases in this sample are missing data in the following areas, listed from those with the most missing cases to those with the le ast: attitudes toward homosexuality, perceptions of inmate credibility, social distance, definitions of rape, punitive orientation, willingness to respond, corruption of authority, counseling role s, male rape myth, and victim blaming. Although it is impossible to determine precisely if data are missing completely at random or missing at random since we do not know the values of the missing data, I assume that the data in this instance are missing at random. It is likely that the true va lues of the variables missing in this data are somewhat related. For example, the results of prior research (Anderson, 2004; Burt & DeMello, 2002; Davies et al ., 2006; Davies & Rogers, 2006; Eigenberg, 2000a; Ford et al., 1998; Mitchell et al., 1999; Wakelin & Long, 2003) a nd this analysis (see Chapter 5) indicate that attitudes toward homosexuality and rape myth s are related to attitude s about victim blaming and perceptions of inmate credibility. For data missing at random, listwise deletion can bias results by produc ing larger standard errors and resulting in under-estimat ed test statistics (A llison, 2002: 6). This can result in a type II error of falsely accepting the nul l hypothesis, and concluding insignificant variables that may, in fact be significant in pred icting the dependent variable (W eisburd, Petrosino, & Mason, 1993). 77

PAGE 78

Though there is various imputation methods that can be used to address data missing at random, the expectation-maximization (EM) method is used to address missing data for the analysis of credibility of victims (see Chapter 5). The reason imputation was used in this part of analysis is that I examine perceptions of credibility separately by sex, which reduced the number of complete cases for females in the sample to less than 100, the size necessary for statistical inference (Groves, Cialdini, & Couper, 1992). Expectation-maximization is a general method for obtaining maximum likelihood estimates when data are missing (Allison, 2002). According to Allison (2002: 13), this method chooses as estimates those values that, if tr ue, would maximize the probability of observing what has, in fact, been observed. The estima tors, including standard errors, that result from maximum likelihood methods are also considered unbiased, efficient, and normal (Agresti & Finlay, 2002; Allison, 2002; Studenmund, 2001). Therefore, because EM uses maximum likeli hood methods that produce relatively efficient estimates, EM imputation methods are used in th e analysis of inmate credibility, otherwise, listwise deletion results are reported. Regression imputation methods, like EM, are also relatively simple to use in SPSS; however, EM is the preferred method as it uses all available variables to predict the values for the missing data and is applicable to a wide pattern of missing data (Allison, 2002; Dempster, Laird, & Rubin, 1977; Little, 1983;). Likewise, because the results of the regression methods are similar to those found using EM, the results reported for the credibility analysis are obtained using the more sophisticated me thod of EM imputation. Perceptions of Professional Orientation and Inmate Attitudes A paired sample t -test is conducted in order to exam ine whether correctional officers expressed attitudes about their profession are si gnificantly different from their perceived attitudes. I compare the mean level of agreement with items that examined their expressed 78

PAGE 79

outlook about their job, their work environment, and inmates with the mean level of agreement that they projected their colleagues to posse ss. This comparison is conducted across all professional orientation items, including attitude s toward counseling roles, punitive orientation, corruption of authority, and social distance from inmates. I c onduct this same means comparison test to determine if officers express victim blaming attitude s and adherence to rape myths different from their perceptions of inmate attitudes. The following chapters present the results of th e analysis. Chapter 4 presents the results of descriptive statistics and includes descriptions of the independent and dependent variables. Chapter 5 reviews the multivariate results examining how correctional officers perceive sexual assault victims in jail by exploring attitudes about victim blaming and credibility of inmates who report rape. Chapter 6 examines definitions of rape and officer willingness to respond to incidents of consensual sex and sexual assault among inmates. Chapter 7 discusses officer views of their fellow officers profe ssional orientation to determine whether their perceptions are accurate. This chapter also compares officer responses to victim blaming and male rape myth items to their perceptions of inmate attitudes ab out these items. Finally, Chapter 8 presents a discussion of the findings of this research, its limitations and contributions, and suggestions for future research. 79

PAGE 80

Table 3-1. Population, sample, and response rate by facility Target population (N = 2822) % Sampled population (N = 1161)a % Sample (N = 376) % Response rate (%)b Alachua Charlotte Clay Escambia Jacksonville Jefferson Leon Palm Beach Pasco Pinellas Sumter Taylor Walton 227 151 118 92 444 13 223 272 295 860 43 33 51 8.0 5.4 4.2 3.3 15.7 0.5 7.9 9.6 10.5 30.5 1.5 1.2 1.8 75 44 47 48 156 8 68 149 87 403 40 18 18 6.5 3.8 4.0 4.1 13.4 0.7 5.9 12.8 7.5 34.7 3.4 1.6 1.6 23 11 15 12 30 5 18 36 40 171 5 8 2 6.1 2.9 4.0 3.2 8.0 1.3 4.8 9.6 10.6 45.5 1.3 2.1 0.5 30.7 25.0 31.9 25.0 19.2 62.5 26.5 24.2 46.0 42.4 12.5 44.4 11.1 a Sampled population differs from the target popula tion as not all officers employed at the jails were working the shifts visited, b The response rate was calculated by dividing the number of surveys returned from the facility by the number of surveys distributed in the facility. Table3-2. Facility characteristics Inmate count Housing type Supervision type County size a Alachua Charlotte Clay Escambia Jacksonville Jefferson Leon Palm Beach Pasco Pinellas Sumter Taylor b Walton 980 529 459 770 2833 75 1161 1993 1300 3128 235 -221 Podular Dormitory, Podular Dormitory, Linear, Podular Dormitory, Linear Podular Podular Podular Dormitory, Linear, Podular Podular Dormitory, Podular Dormitory, Podular -Podular Direct Direct Remote Direct, Remote Remote Remote Direct Direct Direct Direct, Remote Remote -Remote 240,082 152,814 182,023 306,407 849,159 14,451 260,945 1,266,451 462,715 922,893 72,246 19,771 52,881 a United States Census Bureau (2008), b Missing data is due to facility refusal to provide information. 80

PAGE 81

Table 3-3. Jail and sample characteristics Codes Sample (N = 376) % Mean SD Target population (N=2822) % Sex Male 0 260 69.1 1953 69.2 Female 1 115 30.6 869 30.8 Missing 1 0.3 Race and ethnicity a White 0 273 77.9 1836 65.1 Black 1 61 16.2 820 29.1 Other 1 8 2.1 59 2.1 Hispanic 1 20 5.4 107 3.8 Missing 14 3.7 Race b White 0 293 77.9 1943 68.9 Black 1 61 16.2 820 29.1 Other 1 8 2.1 59 2.1 Missing 14 3.7 Education c 3.23 1.31 HS or GED 1 38 10.1 2822 Some college <1 year 2 79 21.0 >1 year college 3 102 27.1 Associates 4 78 20.7 304 10.7 Bachelors 5 71 18.9 160 5.7 Masters 6 6 1.6 7 0.2 Professional/ doctorate 7 1 0.3 0 Missing 1 0.3 Marital status Married 0 241 64.1 Widowed 1 2 0.5 Divorced 1 62 16.5 Separated 1 7 1.9 Never married 1 60 16.0 Missing 1 4 1.1 Age 4.38 1.94 18-25 1 26 6.9 26-30 2 40 10.6 31-35 3 58 15.4 36-40 4 79 21.0 41-45 5 71 18.9 46-50 6 44 11.7 51-55 7 29 7.7 56-60 8 18 4.8 Over 60 9 8 2.1 Missing 3 0.8 81

PAGE 82

Table 3-3. Continued Codes Sample (N = 376) % Mean SD Target population (N = 2822) % Political party Republican 0 155 41.2 Democratic 1 133 35.4 Independent 1 63 16.8 Other 1 17 4.5 Missing 8 2.1 Political orientation 4.44 0.99 Extremely liberal 1 4 1.1 Very liberal 2 7 1.9 Liberal 3 26 6.9 Middle of the road 4 163 43.4 Conservative 5 119 31.6 Very conservative 6 28 7.4 Extremely conservative 7 11 2.9 Missing 18 4.8 Religiosity 5.76 2.43 Not at all religious 1 26 6.9 2 19 5.1 3 29 7.7 4 21 5.6 5 74 19.7 6 39 10.4 7 53 14.1 8 76 20.2 9 15 4.0 Extremely religious 10 19 5.1 Missing 5 1.3 Know prisoner Yes 0 253 67.3 no 1 122 32.4 Missing 1 0.3 Job satisfaction n % 6.92 2.09 Not at all satisfied 1 4 1.1 2 11 0.9 3 16 4.3 4 14 3.7 5 47 12.5 6 39 10.4 7 76 20.3 8 86 22.9 9 44 11.7 Extremely satisfied 10 38 10.1 Missing 1 0.3 82

PAGE 83

83 Table 3-3. Continued Codes Sample (N = 376) % Mean SD Target population (N = 2822) % Job stress 5.96 2.17 No stress 1 7 1.9 2 18 4.8 3 25 6.6 4 35 9.3 5 79 21.0 6 50 13.3 7 52 16.8 8 45 12.0 9 24 6.4 Extreme stress 10 23 6.1 Missing 8 2.1 Job experience d (mode = 2) 10.25 6.99 Four years or less 95 26.0 5-9 years 94 26.0 10-14 years 74 20.2 15-19 years 58 15.8 20-24 years 34 9.3 25 years or more 11 3.0 Missing 10 2.7 Shift Day shift 1 168 44.7 Evening shift 2 70 18.6 Night shift 3 120 31.9 Missing 18 4.8 Two shift responses Evening shift 2 2 0.5 Night shift 3 8 2.1 Facility geography Rural 1 31 8.2 Suburban 2 71 18.9 Urban 3 241 64.1 Missing 33 8.8 a Hispanic is categorized as a separate racial category, b Hispanics included in the White racial category, c Minimum education achieved, d Nearest year .

PAGE 84

Table 3-4. Factor loadings and reliability scores of independent variables N = 376 Min, max a Mean (SD) Cronbachs alpha Factor loading Professional orientation Counseling roles 369 1, 6 3.33 (1.26) .77 Rehabilitation programs should be left to mental health professionals 371 1, 6 3.30 (1.41) .63 Counseling is a job for counselors, not correctional officers 374 1, 6 3.22 (1.55) .92 If a correctional officer wants to do counseling, he should change jobs 375 1, 6 3.47 (1.61) .91 Punitive orientation 361 1.33, 6 3.58 (1.08) .58 Rehabilitation programs are a waste of time and money b 371 1, 6 3.42 (1.44) .75 Improving jails for inmates makes them worse for officers b 368 1, 6 3.15 (1.56) .71 A military regime is the best way of running a jail b 368 1, 6 4.20 (1.41) .76 Corruption of authority 369 2, 6 5.13 (0.79) .74 You cant ever completely trust an inmate b 374 1, 6 5.06 (1.24) .65 A good principle is not to get close to inmates b 373 1, 6 5.36 (1.02) .71 A personal relationship with an inmate invites corruption b 372 1, 6 5.45 (0.94) .75 You must keep conversations with inmates short and businesslike b 372 1, 6 4.80 (1.33) .72 If an officer is lenient with inmates they will take advantage of him/ her b 373 1, 6 4.96 (1.11) .69 Social distance 358 1.4, 6 3.73 (0.88) .67 A correctional officer should work hard to earn trust from inmates 371 1, 6 3.39 (1.54) .55 Its important for a correctional officer to have compassion 371 1, 6 3.03 (1.33) .68 You get to like the inmates you supervise 372 1, 6 4.09 (1.28) .66 Sometimes a correctional officer should be an a dvocate for an inmate 368 1, 6 3.97 (1.33) .64 The way to get respect from inmates is to take an interest in them 371 1, 6 4.21 (1.24) .77 84

PAGE 85

Table 3-4. Continued N = 376 Min, max a Mean (SD) Cronbachs alpha Factor loading Attitudes toward homosexuality (ATH) 338 1, 6 3.13 (1.37) .94 Male homosexual couples should be allo wed to adopt children the same as heterosexual couples b 368 1, 6 2.75 (1.70) .82 I would not be too upset if I learned that my son was a homosexual b 370 1, 6 2.84 (1.64) .77 Homosexual behavior between two men is ju st plain wrong 368 1, 6 2.75 (1.70) .83 Male homosexuality is merely a different kind of lifestyle that should not be condemned b 368 1, 6 3.15 (1.66) .71 Just as in other species, male homosexuality is a natural expression of sexuality in human b 366 1, 6 2.81 (1.55) .80 I think male homosexuals are di sgusting 366 1, 6 3.24 (1.72) .79 Male homosexuals should not be allowed to teach school 360 1, 6 3.91 (1.66) .76 If a man has homosexual feelings, he should do everything he can to overcome them 361 1, 6 3.27 (1.62) .87 Male homosexuality is a perversion 359 1, 6 3.36 (1.69) .88 The idea of male homosexual marriages seems ridiculous to me 363 1, 6 2.81 (1.71) .90 Male rape myth 1, 6 Most men who are raped by a man are very upset by the incident (Myth 1) 372 1, 6 1.96 (1.09) It is impossible for a man to rape a man b (Myth 2) 371 1, 6 1.65 (1.17) Most men who are raped by a man do not need counseling after the incident b (Myth 3) 372 1, 6 1.91 (1.13) Even a big, strong man can be raped by another man (Myth 4) 371 1, 6 2.12 (1.35)a Codes: 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongl y disagree, b Reverse-coded item 85

PAGE 86

Table 3-5. Factor loadi ngs and reliability scores of dependent variables N = 376 Min, max a Mean (SD) Cronbachs alpha Factor loading Victim blaming a 371 1, 6 2.06 (0.97) .93 Inmates who have previously consented to participate in sexual acts in jail get what they deserve if they are raped by other inmates b 374 1, 6 2.33 (1.33) .82 Some inmates deserve to be raped in jail because of the way they act b 375 1, 6 1.96 (1.07) .82 Homosexual inmates get what they deserve if they are raped in jail b 374 1, 6 1.90 (1.02) .90 Inmates who take money or cigarettes in exchange for consensual sexual acts get what they deserve if they are raped by other inmates b 374 1, 6 2.21 (1.26) .89 Inmates who dress or talk in feminine ways get what they deserve if they are raped in jail b 374 1, 6 2.02 (1.14) .90 Most men who are raped by a man are somewhat to blame for not being more careful b 373 1, 6 2.01 (1.09) .79 Most men who are raped by a man are to blame for not escaping or fighting off the man b 374 1, 6 2.08 (1.14) .79 Credibility of victims c An inmate tells you he was raped. Ho w likely are you to believe him if he: 357 1.44, 5 3.55 (0.71) .91 Is a muscular inmate b 364 1, 5 3.34 (1.05) .80 Is a homosexual inmate b 363 1, 5 3.77 (0.88) .71 Is a young inmate b 364 2, 5 4.08 (0.65) .69 Is a gang member b 364 1, 5 3.34 (1.06) .82 Is an inmate who owes money b 364 1, 5 3.76 (0.84) .75 Is drunk or high on drugs b 358 1, 5 3.29 (0.99) .83 Delayed reporting the incident b 362 1, 5 3.53 (0.90) .78 Has previously consented to sex with other inmates b 361 1, 5 3.30 (1.01) .81 Has mental health problems b 361 1, 5 3.48 (0.89) .70 Rape definitions a 359 1, 6 4.81 (1.01) Inmate Jones physically overpowers inmate Smith. Smith has sex with Jones. Smith has been raped. b 363 1, 6 5.16 (0.98) .87 .54 86

PAGE 87

87 Table 3-5. Continued N = 376 Min, max a Mean (SD) Cronbachs alpha F actor l oading Inmate Jones threatens to tell other inma tes that inmate Smith is an informant unless inmate Smith engages in sexual acts. Smith has sex with Jones. Smith has been raped. b 366 1, 6 4.74 (1.33) .89 Inmate Jones tells inmate Smith he will kill him unless Smith has sex with Jones. Smith has sex with J ones. Smith has been raped. b 367 1, 6 5.08 (1.09) .86 Inmate Smith is an informant. Inmates Jones provides protection for Smith but demands that Smith participate in sexual acts. Smith has sex with Jones. Smith has been raped. b 363 1, 6 4.49 (1.40) .86 Inmate Jones loans inmate Smith money or some goods. Smith cannot pay Jones back. Jones tells Smith that he can par ticipate in sexual acts to pay off his debt or inmate Jones will beat inmate Smith severely. Smith has sex with Jones. Smith has been raped. b 364 1, 6 4.58 (1.40) .85 Willing to respond: report and prevent scale a (Respond 1) 364 2.5, 6 5.22 (0.71) .70 Jail officers should encourage inmates to report consensual sexual acts that occur in jail b 366 1, 6 4.60 (1.39) .63 Jail officers should encourage inmates to report sexual assaults that occur in jail b 369 2, 6 5.48 (0.71) .82 Jail officers should do everything they ca n to prevent consensual sexual acts in j ail b 368 1, 6 5.22 (0.99) .78 Jail officers should do everything they can to prevent sexual assaults in jail b 368 2, 6 5.60 (0.62) .83 Willing to respond: talk to inmates scale a (Respond 2) 368 1, 6 4.29 (1.33) .81 Jail officers should talk to inmates about consensual sexual acts to discourage those activities b 368 1, 6 4.18 (1.49) .92 Jail officers should talk to inmates a bout the risk of sexual assault in jail b 368 1, 6 4.41 (1.42) .92 Willing to respond: proactive action scale a (Respond 3) 363 2, 6 4.83 (0.97) .67 Jail officers should use cell assignments to safe-guard inmates from sexual assaul t b 365 1, 6 4.92 (1.04) .87 Jail officers should refer inmates to pr otective custody to safe-guard them from sexual assault b 366 1, 6 4.72 (1.20) .87a Codes: 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongl y disagree, b Reverse-coded item, c Codes: 1 = always, 2 = generally, 3 = sometimes, 4 = rarely, 5 = never

PAGE 88

CHAPTER 4 DESCRIPTIVE STATISTICS Introduction This chapter reviews the descriptive statistics of the independent and dependent variables. What follows is a discussion of the sample re sponses to questions regarding professional orientation, attitudes toward hom osexuality, acceptance of male rape myths, victim blaming, credibility of victims, definitions of rape, and willingness to respond to incidents of sexual assault. This chapter will also explore whether there are significant differences between sex and education sub-samples in response to any of the aforementioned variables. Chapters 5 through 7 will focus on multivariate models and th e results of the quantitative analysis. Independent Variables The following discussion presents means to examine descriptive statistics of the independent variables. Means for the full samp le are presented, as well as means for sex and education sub-samples. The descriptive statistics for the independent variables are presented in Tables 4-1 and 4-2, and indicate both means and standard deviations for individual items and indices. I chose to conduct a separate analysis of response by sex because research indicates sex is an important factor in bl aming attitudes and attitudes to ward homosexuality (Eigenberg, 2000a; Herek, 1988; MaCrae & Shepherd, 1989; Pollar d, 1992). Research also indicates that education can influence attitudes about rape my ths and offenders and rehabilitation (Burt, 1980; Cullen et al., 1993). The education sub-samples c onsist of those with a college degree, including an associates degree and higher, and those with out a college degree. In addition to comparing mean values, t -test results (Tables 4-1 through 4-4) are presented to examine whether there are significant differences in the means of the inde pendent variables between males (n = 260) and females (n = 115) and officers with (n = 156) and without (n = 219) a college degree. 88

PAGE 89

Professional Orientation As discussed in Chapter 3, I constructed an i ndex to measure correctional officer attitudes toward counseling roles as part of their job function. Full sample re sponses to this index indicate that there is moderate support for counseling roles among the Flor ida jail correctional officers who responded to the survey. Responses to item s on this index ranged from 1 to 6 with higher responses indicating more support for counseling roles as part of the correctional officer job. The average mean sample response to the index items is 3.33, with a standard deviation of 1.26 (Table 4-1), indicating overall uncertainty a bout this index. The response option of three indicates a response of not sure, probably agree, while the response value of four indicates not sure, probably disagree. Full sample response means to individual items on this index range from 3.22 (Counseling is a job for counselor s, not correctional officers) to 3.47 (If a correctional officer wants to do counseling, he sh ould change jobs). The average response to this index is higher among males in the sample ( x = 3.36) than females ( x = 3.29). Within the education sub-samples, the average response to this index is higher among those with no degree ( x = 3.40) than those with a degree ( x = 3.25) (Table 4-2). I conduct independent samples t -tests to examine whether differences among sex and education sub-samples were significant for all of the independent indices and individual index items. There is no significant difference between males and females or offi cers with or without a college degree in response to the counseling role s index or any of the i ndividual counseling role items. While the sex findings contrast previous re search that indicates females as more favorable to the human service component of correctional work (Farkas, 1999; Griffin, 2002; Jurik, 1985; Kifer et al., 2003; Pogrebin & Poole, 1998), the education analysis suppo rts the inconsistent results of prior research (Farkas, 200 1; Jurik, 1985; Kassebaum et al., 1964). 89

PAGE 90

As a whole, correctional officers are somewh at punitive and average 3.58 on the items in the punitive orientation index (Table 4-1). Again, responses to ite ms on this index range from a 1 to 6 with higher values indicating more punitive attitudes (items were reverse-coded so that a response value of three indicated, not sure, probably disagree while a value of four represented not sure, probably agree). The item producing the least punitive response among the full sample states that improving jails for inmates makes them worse for officers ( x = 3.15). The item that garnered the most punitive response stat ed that a military regime is the best way of running a jail ( x = 4.20). Males have a slightly greater mean ( x = 3.63) on this index than females ( x = 3.49) indicating that male respondents tend to be slightly more punitive than female respondents are. The t -test results indicate no significant difference between males and females in response to the punitive orientation index or individual punitive orientation items. There are no substantive or significant differences between those with or without a college degree in response to this index or its individual items (Table 4-2). Female respondents show greater concern over the corruption of their authority ( x = 5.17) than male respondents ( x = 5.11) (Table 4-1). Response values for this index also range from 1 to 6 with higher values representing greater c oncern over corruption of authority (items were reverse-coded so that a response value of 5 indi cates, agree while a response of six indicates, strongly agree). The full sample mean ( x = 5.13) for the corruption of authority index lies between the female and male mean, and indicates that the sample is rather concerned with the corruption of their authority by inmates, whic h reflects findings from previous research (Eigenberg, 2000a; Toch & Klofas, 1982). The t -tests reveal significant differences betw een males and females in response to two individual corruption of authority items, similar to previous studies (Eigenberg, 2000a; Farkas, 90

PAGE 91

1999). Male mean responses to the statement a good principle is not to ge t close to inmates indicate that males have less agreement with the item and less concern over closeness or corruption of authority than females (Table 4-31. Female responses to this item may reflect the difficulties presented by cross-gendered correctional supervision that may not be present for same-sex correctional supervis ion (Alpert & Crouch, 1991). Males have significantly higher responses than females to the statement if an o fficer is lenient with in mates they will take advantage of him/ her. Male responses to th is item are also lower than female responses, indicating less agreement with the item and less concern over lenien cy and corruption of authority than females (Table 4-3). Although officers with and without a college ed ucation vary somewhat in their responses to individual items in the corruption of authority index, there are no substantive or significant differences in average responses to the index (T able 4-2). I am aware of only one prior study that indicates that education (s pecifically, lower education) pred icts concern over corruption of authority (Eigenberg, 2000a). The item that produced the highest concern over corruption of authority among the full sample states, a pers onal relationship with an inmate invites corruption ( x = 5.45). The item that produced the leas t concern over corruption of authority stated, officers must keep conversations with inmates short and businesslike. These results indicate that officers were more concerned w ith the effect of pers onal relationships over conversation on their ability to effectively manage inmates and exert their authority. The full sample shows some preference for social distance from inmates ( x = 3.73) (Table 4-1), similar to prior studies (Eigenberg, 2000a; Farkas, 1999; Kassebaum et al., 1964; Shamir & Drory, 1981; Toch & Klofas, 1982). Again, respons es on this index range from 1 to 6 with higher values representing greater preference for social distance. The item generating the most 91

PAGE 92

preference for social distance among the full sample states, the way to get respect from inmates is to take an interest in them ( x = 4.21). The item with the lowest sample mean states that it is important for a correctional officer to have compassion ( x = 3.03). Regarding simple means comparison, females and those with no college degree show somewhat more preference for social distance ( x = 3.76 and 3.77, respectively) than males and officers with a college degree ( x = 3.72 and 3.68, respectively) (Tables 4-1 and 4-2). The t -test analysis reveals no significant differences between the sex or e ducation sub-samples on any social distance item or the social distance index (Tables 4-1 and 4-2). Overall, the results of the desc riptive analysis thus far indica te that female respondents are less professionally oriented than male respondents are. Females in the sample have overall less support for counseling roles, greater punitive or ientation, greater con cern over corruption of authority, and more preference for social dist ance. With the exception of concern over corruption of authority, respondents in the full sample are somewhat moderate in their attitudes about professional orientation items. The t -test results, however, in dicate that the only significant differences between the sexes in resp onse to the professional orientation index s is that females show more concern over two corrup tion of authority items than males do. Females are significantly more likely to be concerned with whether gett ing close to or having a personal relationship with inmates threatens their author ity. This finding mirrors prior studies that indicate that sex may be importa nt predictors of con cern over corruption of authority (Eigenberg, 2000a; Farkas, 1999). The descriptive analysis reveals no consistent difference between correctional officers with and those without a college degree. Within the education sub-samples, officers without a college degree show more support for counseling roles and more preference for social distance between 92

PAGE 93

themselves and inmates. Those with a colleg e degree have overall le ss support for counseling roles and less preference for social distance betw een themselves and inmates. This result could indicate that some officers w ith a college degree believe that a job distinction should exist between security and treatment staff members. In addition, differences in the social distance items indicate that officers with a degree may be more comfortable having compassion for and liking inmates as well as getting respect by taking an interest in them. There are no notable differences in means between these two s ub-samples regarding punitive orientation and corruption of authority. Additiona lly, there are no significant mean differences in responses to any individual professional orientation item or index. Attitudes toward Homosexuality The full samples mean response to the at titudes toward homosexuality index is 3.13, indicating moderate acceptance of homosexuals and homosexual lifestyles (Table 4-1). Response values for these items ranged from 1 to 6 with higher values representing more acceptance of homosexuality. Four of the te n attitudes toward homosexuality items were reverse-coded (Table 4-1), so that as read they represent opposition to homosexuality. Therefore, response values of three generally indicate, not sure, but probably agree in opposition to homosexuality, where a response of f our would indicate, not sure, but probably disagree with opposition to this lifestyle. In addition, where a va lue of two would indicate the respondent agrees in opposing this lifestyle, a response value of five would indicate more acceptance of homosexuality (disagreement with opposition to the lifestyle). The mean for female respondents on this index is much higher ( x = 4.19) than the mean for male respondents ( x = 2.68) indicating that female respondents ha ve greater tolerance for homosexuals and homosexual lifestyles. The full sample respon ses on the index items ranges from 2.75 (male homosexual couples should be allowed to adopt children) to 3.91 (male homosexuals should 93

PAGE 94

not be allowed to teach school). Ttests indicate that there are significant difference between sexes on the attitudes toward homosexuality in dex and all items on the index (Table 4-1). Specifically, females have significantly highe r means and are thus more tolerant of homosexuality and homosexuals, which supports previous research (Eigenberg, 2000a; Herek, 1988). The descriptive analysis indica tes that officers with a college degree have higher means for this index and most of the items on the index and therefore more tolerant of homosexuality (Table 4-2). However, the t -test results for those with varyin g levels of education reveal no significant differences between officer with or w ithout a college degree on these items (Table 42). The results are not consistent with previous researchs find ing that less educated officers hold more negative attitudes toward bisexual and homosexual lifestyles (Eigenberg, 2000a; Herek, 2002), and that higher education positively affects perceptions of homosexuals (Ohlander, Batalova, & Treas, 2005). Male Rape Myths There are few differences in simple mean comparisons between males and females with regard to attitudes about male rape myths, however. For the two rape myths that, as read, represent statements that males are capable of succumbing to rape, the response code of one indicates strongly agree and the code of six re presents strongly disagree. The two rape myths that are reverse coded, however, are statements that, as read, indicate that it is not possible for males to be raped. These two items are re verse-coded so that one represents strongly disagree and six represents strongly agree. Ther efore, male rape myth responses range from 1 to 6 with higher values representing greater adherence to male rape myths. Males show slightly greater acceptance of 3 out of 4 of the male rape myth items (it is impossible for a man to rape a man, most raped men do not need counseling af ter the incident, and even a big, strong man 94

PAGE 95

can be raped by another man) (Table 4-1). The rape myth receiving the most support (disagreement with the statement and adherence to the rape myth) among the full sample stated that even a big, strong man can be raped by another man ( x = 2.12). The myth receiving the least support (reverse coded so that higher scores reflect adherence to the rape myth) stated that it is impossible for a man to rape a man ( x = 1.65). T -tests reveal that males have significantly higher levels of adhere nce to the male rape myth that males who are raped by a man do not need counseling after the incident (Table 4-1), similar to other studies (Feild, 1978; Jenkins & Dambrot, 1987; Kopper, 1996). This indicates that males tend to believe that strong men do not need counse ling after traumatic events. The differences between officers with and officers without a college de gree are neither significant nor consistent. Officers with a college degree have higher means and more adherence to two of the four rape myth items (most raped men are ve ry upset by the incident and even a big, strong man can be raped by another man). None of the differences between these groups of officers is large or significant, however (Table 4-2). Dependent Variables This section presents means in order to e xplore descriptive statis tics of the dependent variables in this analysis. I will discuss means for the overall sample, as well as means for the sex and education sub-samples. Following this, I will present t -test results to examine whether there are significant differences in the mean s of the dependent variables across sex and education. The descriptive statis tics for the dependent variables are presented in Tables 4-3 and 4-4, and indicate both means and standard deviations for the individual items and indices. For each item and index, the means and standard deviations for the full sample and the sex and education sub-samples are presented. 95

PAGE 96

Victim Blaming Overall, the full sample exhibits rela tively low levels of victim blaming ( x = 2.07, reversecoded so that a value of two indicates, disagr ee) (Table 4-3). Response values on the victim blaming index ranged from 1 to 6 with higher responses representing greater blaming attitudes. The item with the highest mean and level of victim blaming am ong the full sample states that inmates who have previously consented to sexual acts get what they deserve if they are raped by other inmates ( x = 2.34). The item with the lowest mean or level of victim blaming states that homosexual inmates get what they deserve if they are raped in jail ( x = 1.90). Across the victim blaming index and all indi vidual index items, male respondents exhibit greater tendencies to victim blame than fema le respondents do (Table 4-3). Likewise, the t -test results indicate that males have significantly high er means than females in response to the victim blaming index and all individual items on the index (Table 4-3). These re sults support previous research indicating that males are more like ly than females to blame victims (MaCrae & Shepherd, 1989; Pollard, 1992). The descriptive an alysis reveals that across all victim-blaming items and the victim blaming index, officers with a college degree have higher levels of victim blaming (Table 4-4). There are no significant di fferences between these groups of officers with respect to victim blaming, however. Inmate Credibility Though there were some differences among ma les and females for individual items measuring the credibility of certa in inmates who report rape, the simple mean responses to the credibility of inmates index is re latively similar between males ( x = 3.46) and females ( x = 3.74) (Table 4-3). Credibility response options range from 1 to 5, with higher values representing greater inmate credibility. These items were also reverse-coded so that a response of one represented a tendency to never believe the in mate, a three indicated that the respondent 96

PAGE 97

sometimes believed the inmate, and a response of five indicated that they would always believe the inmate. The full sample mean lies between that of males and females ( x = 3.55). Overall, the sample appears to give credibility to inmates who report rape The individual items that garner the lowest means indicate that inmates who are hi gh or drunk and those who have previously consented to sex w ith other inmates are given the least amount of credibility ( x = 3.29 and 3.30, respectively). The items that received the higher means indicate that among all types of inmates represented, homosexual inmates or inmates who owe money receive the most credibility ( x = 3.77 and 3.76, respectively). Both this finding and the victim blaming results indicate that blaming attitudes a nd lack of credibility are not au tomatic responses to inmates who are homosexuals. The t -test analysis reveals that ma les have significantly lower means for the credibility of inmates index and five individual items in the i ndex (Table 4-3). Males show significantly less belief in the credibility of in mates who are muscular, homosexual, members of a gang, in debt, and those who have previously consented to sex with other inmates. There are no consistent differences between officers with a college degr ee and officers without a college degree across inmate credibility index or its individual items (Table 4-4). Officers with a college degree have slightly higher means in response to the homos exual inmate credibility item. Officers have identical average responses to the items dea ling with drunk or high inmates and inmates who delay reporting. Officers with no college degree have higher levels of agreement on all other credibility items. However, there are no signifi cant differences in means between these groups of officers. Rape Definitions In defining what constitutes rape, male and female respondents have similar levels of agreement, although males have slightly lower le vels of agreement on the definition of rape 97

PAGE 98

index ( x = 4.77 and 4.90, respectively) (Table 4-3). This finding is similar to Eigenbergs (2000a) as it did not find signifi cant differences across sex on this variable. The response values for definition of rape items range from 1 to 6 with higher values representing agreement in defining the incident as rape. As stated, agreement with each statement would indicate belief that the inmate was raped. However, these items are reverse-coded so that a response of five indicated a response of agree that the incident represents rape and a response of four indicated, not sure, but probably agree. The full sample mean for the definition inde x is 4.81, which lies in between male and female respondent means, and indicates moderately liberal attitudes about what constitutes rape. The items that produce greater tendencies to define the incident as rape included incidents where the offender physically overpowered or threatened to kill the victim ( x = 5.16 and 5.08, respectively). The item that produc ed least tendency to define the situation as rape involved the offender threatening to disclose the victim as an informant unless they submit to sexual acts ( x = 4.49). These findings indicate that respondents may be more likely to define situations as rape when they involve threats or physical assaults. The t -tests reveal no signi ficant differences between definitions of rape across sex, however. Within the e ducation sub-samples, officers with a degree have higher levels of agreement w ith rape definition index and four of its five individual items (Table 4-4). The differences between the groups are neither substantial nor significant, however. Willingness to Respond There are similarities between male and female respondents willingness to respond to rape or sexual assault incidents, and these similarities persisted across the thre e indices (willingness to encourage inmate reporting and officer preven tion, willingness to talk to inmates, and willingness to engage proactive measures). Respondents in the full sample average 5.22 on the 98

PAGE 99

index that measured their willingness to respond by encouraging inmate reporting and officer prevention (Table 4-3). Response options range from 1 to 6 with higher values indicating greater willingness to respond as indicated. As stated, agreement with each statement would indicate belief that officers should adopt th at response. However, these items are also reverse-coded so that a response of five indicates, agree and a re sponse of four indicates, not sure, but probably agree. Officers have a lower mean on the index that measured willingness to respond by talking to inmates about consensual sex and about the risks of sexual assault ( x = 4.29), however, their mean response to the proactive measures inde x lies between the ot her response indices ( x = 4.83). Overall, respondents are le ast willing to respond by talking to inmates about consensual sexual acts to discourag e those activities ( x = 4.18) and are most willing to respond by doing everything they can to prevent sexual assaults ( x = 5.60). There are no significant differences between ma les and females in response to any of the willingness to respond indices or individual ite ms on these indices, similar to findings by Eigenberg (1994) (Table 4-3). There are greater mean differences between the education subsamples than the sex sub-samples. Response m eans differ significantly acr oss level of education for two individual items and one index. Officers with no college degree have higher means for the individual item that stated jail officers s hould do everything they can to prevent consensual sexual acts in jail (Table 4-4). Officers with a college degree, however, have significantly higher response means to the index that meas ured willingness to respond by acting proactively and the item in this index that referred to the use of cell assignments to safeguard inmates (Table 4-4). Conclusion This chapter reveals that officers are concerned with many aspects of their profession as well as instances of sexual victimization among inmate s in jails. The descriptive analysis of their 99

PAGE 100

professional orientations reveal s that they have moderate s upport for counseling roles, are somewhat punitive, are concerned with the corrup tion of their authority, and somewhat prefer social distance in their relationships with inmate s. Regarding professional orientation toward their jobs, female officers are significantly more likely to report concern over the corruption of their authority than male officers are. Although the full sample reveals moderate acceptance of homosexuality, female respondents have significantly higher levels of acceptance than males on all items in this category. Male officers also have significantly greater acceptance of one of the male rape myth beliefs. The full sample also displays relativel y low levels of victim blaming, although males have significantly higher levels of victim blaming attitudes than females. Although the full sample gives overall high credibility to inmate s who report rape, males have significantly lower means in response to the credibil ity of specific types of inmates who report rape, indicating that females often see these inmate reports as more credible. The simple means comparison indicates that respondents more readily define some incidents as rape, although mean sample response s to the definition of rape index indicates a tendency to agree that the vignett es describe instances of rape. In addition, there are no significant differences across sub-sa mples in response to any of these items. Officers in the sample also reveal a willingness to respond to inst ances of rape and sexual assault, although they favor some responses over others. Although ther e are no significant differences between sexes in the willingness to respond items, there are some differences between officers with and without a college degree in response to these indices. Specifically, officers without a college degree are significantly more likely to ag ree that officers should do ever ything they can to prevent consensual sexual acts in jail. Officers with a college degree are more likely to agree that 100

PAGE 101

101 officers should use proactive measures (like cell assignments) to safeguard inmates from sexual assault. The next two chapters will examine th e relationships between these variables and will examine how multivariate models react to these data.

PAGE 102

Table 4-1. Descri p tive statistics b y sex: inde p endentvariables N = 376 (male, female) Min, max a Full sample mean (SD) Male n = 260 mean (SD) Female n = 115 mean (SD) Mean difference c Professional orientation Counseling roles 371 ( 256 114 ) 1, 6 3.33 ( 1.26 ) 3.36 ( 1.26 ) 3.29 ( 1.25 ) .08 Rehabilitation programs should be left to mental health professionals 372 (256, 115) 1, 6 3.30 (1.41) 3.32 (1.44) 3.28 (1.33) .04 Counseling is a job for counselors, not correctional officers 376 ( 260 115 ) 1, 6 3.22 ( 1.55 ) 3.24 ( 1.56 ) 3.19 ( 1.54 ) .05 If a correctional officer wants to do counseling, he should change jobs 375 (260, 114) 1, 6 3.47 (1.61) 3.52 (1.59) 3.39 (1.65) .14 Punitive orientation 364 ( 250 113 ) 1.33, 6 3.58 ( 1.08 ) 3.63 ( 1.09 ) 3.49 ( 1.06 ) .13 Rehabilitation programs are a waste of time and money b 372 ( 257 114 ) 1, 6 3.42 ( 1.44 ) 3.49 ( 1.43 ) 3.24 ( 1.43 ) .24 102 Improving jails for inmates makes them worse for officers b 370 ( 255 114 ) 1, 6 3.15 ( 1.56 ) 3.20 ( 1.55 ) 3.04 ( 1.57 ) .15 A military regime is the best way of running a jail b 370 ( 255 114 ) 1, 6 4.20 ( 1.41 ) 4.22 ( 1.41 ) 4.20 ( 1.37 ) .01 Corruption of authority 369 ( 254 114 ) 1, 5 5.13 ( 0.79 ) 5.11 ( 0.82 ) 5.17 ( 0.73 ) -.06 You cant ever completely trust an inmate b 374 ( 258 115 ) 1, 6 5.06 ( 1.24 ) 5.05 ( 1.27 ) 5.09 ( 1.18 ) -.04 A good principle is not to get close to inmates b 373 ( 257 115 ) 1, 6 5.36 ( 1.02 ) 5.29 ( 1.09 ) 5.52 ( 0.84 ) -.22* A personal relationship with an inmate invites corruption b 372 ( 256 115 ) 1, 6 5.45 ( 0.94 ) 5.39 ( 1.01 ) 5.59 ( 0.77 ) -.20* You must keep conversations with inmates short and businesslike b 372 (257, 114) 1, 6 4.80 (1.33) 4.77 (1.36) 4.88 (1.26) -.10 If an officer is lenient with inmates they will take advantage of him/ her b 373 (257, 115) 1, 6 4.96 (1.11) 5.04 (1.06) 4.78 (1.22) .26*

PAGE 103

Table 4-1. Continue d N = 376 (male, female) Min, max a Full sample mean (SD) Male n = 260 mean (SD) Female n = 115 mean (SD) Mean difference c Social distance 358 ( 244 113 ) 1, 5.6 3.73 ( 0.88 ) 3.72 ( 0.89 ) 3.76 ( 0.89 ) -.04 A correctional officer should work hard to earn trust from inmates 371 ( 255 115 ) 1, 6 3.39 ( 1.54 ) 3.22 ( 1.54 ) 3.75 ( 1.48 ) -.54 Its important for a correctional officer to have compassion 371 ( 256 114 ) 1, 6 3.03 ( 1.33 ) 3.06 ( 1.32 ) 2.96 ( 1.36 ) .10 You get to like the inmates you supervise 372 ( 256 115 ) 1, 6 4.09 ( 1.28 ) 4.18 ( 1.27 ) 3.89 ( 1.26 ) .28 Sometimes a correctional officer should be an advocate for an inmate 368 (253, 114) 1, 6 3.97 (1.33) 3.98 (1.34) 3.94 (1.28) .05 The way to get respect from inmates is to take an interest in them 371 ( 255 115 ) 1, 6 4.21 ( 1.24 ) 4.18 ( 1.26 ) 4.28 ( 1.20 ) -.09 103 Attitudes toward homosexuality (ATH) 341 ( 238 102 ) 1, 6 3.13 ( 1.37 ) 2.68 ( 1.14 ) 4.19 ( 1.27 ) -1.34** Male homosexual couples should be allowed to adopt children the same as heterosexual couples b 369 (254, 114) 1, 6 2.75 (1.69) 2.33 (1.46) 3.70 (1.81) -1.33** I would not be too upset if I learned that my son was a homosexual b 371 (256, 114) 1, 6 2.84 (1.64) 2.42 (1.42) 3.79 (1.73) -1.36** Homosexual behavior between two men is just plain wrong 368 ( 254 113 ) 1, 6 2.75 ( 1.70 ) 2.27 ( 1.43 ) 3.85 ( 1.76 ) -1.58** Male homosexuality is merely a different kind of lifestyle that should not be condemned b 368 (254, 113) 1, 6 3.15 (1.66) 2.88 (1.56) 3.73 (1.73) -.86** Just as in other species, male homosexuality is a natural expression of sexuality in human men b 366 (254, 111) 1, 6 2.81 (1.55) 2.46 (1.41) 3.60 (1.54) -1.12** I think male homosexuals are disgusting 366 ( 254 111 ) 1, 6 3.24 ( 1.72 ) 2.78 ( 1.57 ) 4.30 ( 1.59 ) -1.48**

PAGE 104

Table 4-1. Continue d N = 376 (male, female) Min, max a Full sample mean (SD) Male n = 260 mean (SD) Female n = 115 mean (SD) Mean difference c Male homosexuals should not be allowed to teach school 360 ( 251, 108 ) 1, 6 3.91 ( 1.66 ) 3.52 ( 1.64 ) 4.85 ( 1.27 ) -1.28** If a man has homosexual feelings, he should do everything he can to overcome them 361 (250, 110) 1, 6 3.27 (1.62) 2.87 (1.48) 4.21 (1.54) -1.31** Male homosexuality is a perversion 360 (249, 110) 1, 6 3.37 (1.69) 2.96 (1.57) 4.33 (1.56) -1.34** The idea of male homosexual marriages seems ridiculous to me 365 (252, 112) 1, 6 2.80 (1.71) 2.28 (1.36) 4.00 (1.81) -1.71** Male rape myth 1, 6 Most men who are raped by a man are very upset by the incident (Myth 1) 372 (256, 115) 1, 6 1.96 (1.09) 1.94 (1.03) 1.97 (1.20) .01 104 It is impossible for a man to rape a man b (Myth 2) 371 (256, 114) 1, 6 1.65 (1.17) 1.72 (1.18) 1.48 (1.13) .25 Most men who are raped by a man do not need counseling after the incident b (Myth 3) 372 (256, 115) 1, 6 1.90 (1.12) 1.99 (1.16) 1.70 (1.00) .28* Even a big, strong man can be raped by another man (Myth 4) 371 (255, 115) 1, 6 2.12 (1.35) 2.12 (1.30) 2.10 (1.46) .01 a Codes: 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongl y disagree, b Reverse-coded item, c Mean difference tested through independent samples t -test, p <.05, ** p <.01

PAGE 105

Table 4-2. Descriptive statistics by education:independent variables N = 376 (none, degree) Min, max a Full sample mean (SD) N o d egree n = 219 m ean ( SD ) Degree n = 156 mean (SD) Mean difference c Professional orientation Counseling roles 371 ( 216 154 ) 1, 6 3.33 ( 1.26 ) 3.40 ( 1.24 ) 3.25 ( 1.29 ) .14 Rehabilitation programs should be left to mental health professionals 372 (217, 154) 1, 6 3.30 (1.41) 3.29 (1.43) 3.32 (1.38) -.05 Counseling is a job for counselors, not correctional officers 376 ( 219 156 ) 1, 6 3.22 ( 1.55 ) 3.30 ( 1.55 ) 3.12 ( 1.55 ) .18 If a correctional officer wants to do counseling, he should change jobs 375 (218, 156) 1, 6 3.47 (1.61) 3.61 (1.58) 3.30 (1.62) .30 Punitive orientation 364 ( 214 149 ) 1.33, 6 3.58 ( 1.08 ) 3.57 ( 1.07 ) 3.60 ( 1.11 ) -.03 105 Rehabilitation programs are a waste of time and money b 372 ( 216 155 ) 1, 6 3.42 ( 1.44 ) 3.40 ( 1.46 ) 3.43 ( 1.40 ) -.04 Improving jails for inmates makes them worse for officers b 370 ( 218 151 ) 1, 6 3.15 ( 1.56 ) 3.15 ( 1.53 ) 3.15 ( 1.61 ) .01 A military regime is the best way of running a jail b 370 ( 216 153 ) 1, 6 4.20 ( 1.41 ) 4.19 ( 1.39 ) 4.24 ( 1.41 ) -.07 Corruption of authority 369 ( 216 152 ) 1, 5 5.13 ( 0.79 ) 5.13 ( 0.75 ) 5.13 ( 0.86 ) -.00 You cant ever completely trust an inmate b 374 ( 219 154 ) 1, 6 5.06 ( 1.24 ) 5.16 ( 1.15 ) 4.94 ( 1.36 ) .23 A good principle is not to get close to inmates b 373 ( 219 153 ) 1, 6 5.36 ( 1.02 ) 5.36 ( 1.00 ) 5.37 ( 1.06 ) -.02 A personal relationship with an inmate invites corruption b 372 ( 218 153 ) 1, 6 5.45 ( 0.94 ) 5.47 ( 0.88 ) 5.42 ( 1.03 ) .04 You must keep conversations with inmates short and businesslike b 372 ( 217 154 ) 1, 6 4.80 ( 1.33 ) 4.74 ( 1.35 ) 4.89 ( 1.30 ) -.16 If an officer is lenient with inmates they will take advantage of him/ her b 373 (218, 154) 1, 6 4.96 (1.11) 4.91 (1.13) 5.03 (1.10) -.11

PAGE 106

Table 4-2. Continue d N = 376 (none, degree) Min, max a Full sample mean (SD) No degree n = 219 mean (SD) Degree n = 156 mean (SD) Mean difference c Social distance 358 ( 208, 149 ) 1, 5.6 3.73 ( 0.88 ) 3.77 ( 0.84 ) 3.68 ( 0.94 ) .11 A correctional officer should work hard to earn trust from inmates 371 (216, 154) 1, 6 3.39 (1.54) 3.42 (1.50) 3.34 (1.60) .10 Its important for a correctional officer to have compassion 371 (217, 153) 1, 6 3.03 (1.33) 3.09 (1.32) 2.95 (1.34) .13 You get to like the inmates you supervise 372 (217, 154) 1, 6 4.09 (1.28) 4.15 (1.25) 4.00 (1.30) .16 Sometimes a correctional officer should be an advocate for an inmate 368 (214, 153) 1, 6 3.97 (1.33) 3.94 (1.30) 3.99 (1.35) -.05 The way to get respect from inmates is to take an interest in them 371 (216, 154) 1, 6 4.21 (1.24) 4.29 (1.20) 4.10 (1.30) .22 106 Attitudes toward homosexuality (ATH) 341 (194, 146) 1, 6 3.13 (1.37) 3.09 (1.33) 3.19 (1.41) -.15 Male homosexual couples should be allowed to adopt children the same as heterosexual couples b 369 (215, 153) 1, 6 2.75 (1.69) 2.66 (1.65) 2.89 (1.76) -.26 I would not be too upset if I learned that my son was a homosexual b 371 (216, 154) 1, 6 2.84 (1.64) 2.81 (1.61) 2.89 (1.69) -.09 Homosexual behavior between two men is just plain wrong 368 (214, 153) 1, 6 2.75 (1.70) 2.61 (1.61) 2.96 (1.82) -.35 Male homosexuality is merely a different kind of lifestyle that should not be condemned b 368 (213, 154) 1, 6 3.15 (1.66) 3.20 (1.65) 3.06 (1.68) .11 Just as in other species, male homosexuality is a natural expression of sexuality in human men b 366 (211, 154) 1, 6 2.82 (1.55) 2.72 (1.49) 2.94 (1.62) -.22 I think male homosexuals are disgusting 366 (211, 154) 1, 6 3.24 (1.72) 3.15 (1.74) 3.37 (1.70) -.21 Male homosexuals should not be allowed to teach school 360 (208, 151) 1, 6 3.91 (1.66) 3.86 (1.69) 4.00 (1.60) -.13

PAGE 107

Table 4-2. Continue d N = 376 (none, degree) Min, max a Full sample mean (SD) No degree n = 219 mean ( SD ) Degree n = 156 mean (SD) Mean difference c If a man has homosexual feelings, he should do everything he can to overcome them 361 (207, 153) 1, 6 3.27 (1.62) 3.26 (1.60) 3.30 (1.64) -.04 Male homosexuality is a perversion 360 (207, 152) 1, 6 3.37 (1.69) 3.34 (1.64) 3.42 (1.75) -.09 The idea of male homosexual marriages seems ridiculous to me 365 (211, 153) 1, 6 2.80 (1.71) 2.71 (1.66) 2.94 (1.78) -.20 Male rape myth 1, 6 Most men who are raped by a man are very upset by the incident (Myth 1) 372 (216, 155) 1, 6 1.96 (1.09) 1.90 (1.05) 2.02 (1.13) -.14 It is impossible for a man to rape a man b (Myth 2) 371 (215, 155) 1, 6 1.65 (1.17) 1.74 (1.28) 1.52 (0.99) .22 107 Most men who are raped by a man do not need counseling after the incident b (Myth 3) 372 (217, 154) 1, 6 1.90 (1.12) 2.00 (1.17) 1.77 (1.02) .21 Even a big, strong man can be raped by another man (Myth 4) 371 (216, 154) 1, 6 2.12 (1.35) 2.08 (1.32) 2.15 (1.39) -.05a Codes: 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongl y disagree, b Reverse-coded item, c Mean difference tested through independent samples t -test, p <.05, ** p <.01

PAGE 108

Table 4-3. Descriptive statisti cs by sex: dependent variables N = 376 (male, female) Min, max a Full sample mean (SD) Males n = 260 mean (SD) Females n = 115 mean (SD) Mean differenced Victim blaming a 374 (258, 115) 1, 6 2.07 (0.97) 2.21 (1.03) 1.75 (0.71) .47** Inmates who have previously con sented to participate in sexual acts in jail get what they deserve if they are raped by other inmates b 375 (259, 115) 1, 6 2.34 (1.34) 2.48 (1.41) 2.03 (1.10) .45** Some inmates deserve to be raped in jail because of the way they act b 375 (259, 115) 1, 6 1.96 (1.07) 2.03 (1.11) 1.78 (0.93) .25* Homosexual inmates get what they deserve if they are raped in jail b 375 (259, 115) 1, 6 1.90 (1.02) 2.05 (1.13) 1.57 (0.61) .49** Inmates who take money or cigarettes in exchange for consensual sexual acts get what they deserve if they are raped by other inmates b 374 (258, 115) 1, 6 2.21 (1.26) 2.37 (1.35) 1.85 (0.95) .52** 108 Inmates who dress or talk in feminine ways get what they deserve if they are raped in jail b 374 (258, 115) 1, 6 2.01 (1.14) 2.19 (1.24) 1.62 (0.71) .58** Most men who are raped by a man are somewhat to blame for not being more careful b 374 (258, 114) 1, 6 2.01 (1.08) 2.15 (1.15) 1.70 (0.86) .45** Most men who are raped by a man are to blame for not escaping or fighting off the man b 374 (258, 115) 1, 6 2.08 (1.12) 2.24 (1.96) 1.70 (0.84) .54** Credibility of victims c An inmate tells you he was raped. How likely are you to believe him if he: 358 (243, 114) 1.44, 5 3.55 (0.71) 3.46 (0.70) 3.74 (0.69) -.28** Is a muscular inmate b 364 ( 249 114 ) 1, 5 3.34 ( 1.05 ) 3.14 ( 1.05 ) 3.75 ( 0.93 ) -.60** Is a homosexual inmate b 364 ( 249 114 ) 1, 5 3.77 ( 0.89 ) 3.62 ( 0.92 ) 4.08 ( 0.72 ) -.45** Is a young inmate b 364 ( 249 114 ) 2, 5 4.08 ( 0.65 ) 4.04 ( 0.66 ) 4.18 ( 0.64 ) -.14 Is a gang member b 364 ( 249 114 ) 1, 5 3.34 ( 1.06 ) 3.21 ( 1.05 ) 3.62 ( 1.03 ) -.40**

PAGE 109

Table 4-3. Continued N = 376 (male, female) Min, max a Full sample mean (SD) Males n = 260 mean (SD) Females n = 115 mean (SD) Mean differenced Is an inmate who owes money b 364 ( 249 114 ) 1, 5 3.76 ( 0.84 ) 3.67 ( 0.84 ) 3.97 ( 0.78 ) -.29** Is drunk or high on drugs b 359 ( 244 114 ) 1, 5 3.29 ( 0.99 ) 3.23 ( 0.99 ) 3.42 ( 0.99 ) -.19 Delayed reporting the incident b 362 ( 247 114 ) 1, 5 3.53 ( 0.90 ) 3.50 ( 0.88 ) 3.60 ( 0.93 ) -.09 Has previously consented to sex with other inmates b 361 ( 246 114 ) 1, 5 3.30 ( 1.01 ) 3.22 ( 1.03 ) 3.46 ( 0.95 ) -.24* Has mental health problems b 362 ( 247 114 ) 1, 5 3.48 ( 0.88 ) 3.45 ( 0.89 ) 3.58 ( 0.87 ) -.13 Rape definitions a 359 ( 246 112 ) 1, 6 4.81 ( 1.01 ) 4.77 ( 1.04 ) 4.90 ( 0.95 ) -.13 Inmate Jones physically overpowers inmate Smith. Smith has sex with Jones. Smith has been raped. b 363 (249, 113) 1, 6 5.16 (0.98) 5.17 (0.97) 5.13 (1.01) .04 109 Inmate Jones threatens to tell other inmates that inmate Smith is an informant unless inmate Smith engages in sexual acts. Smith has sex with Jones. Smith has been raped. b 366 (251, 114) 1, 6 4.74 (1.33) 4.70 (1.37) 4.82 (1.26) -.14 Inmate Jones tells inmate Smith he will kill him unless Smith has sex with Jones. Smith has sex with Jones. Smith has been raped. b 367 (252, 114) 1, 6 5.08 (1.09) 5.03 (1.17) 5.19 (0.89) -.17 Inmate Smith is an informant. Inmates Jones provides protection for Smith but demands that Smith participate in sexual acts. Smith has sex with Jones. Smith has been raped. b 363 (249, 113) 1, 6 4.49 (1.40) 4.43 (1.45) 4.63 (1.28) -.22 Inmate Jones loans inmate Smith money or some goods. Smith cannot pay Jones back. Jones tells Smith that he can participate in sexual acts to pay off his debt or inmate Jones will beat inmate Smith severely. Smith has sex with Jones. Smith has been raped. b 364 (250, 113) 1, 6 4.58 (1.40) 4.54 (1.42) 4.67 (1.35) -.15

PAGE 110

Table 4-3. Continue d N = 376 (male, female) Min, max a Full sample mean (SD) Males n = 260 mean (SD) Females n = 115 mean (SD) Mean difference d Willing to respond: report and prevent scale (Respond 1) a 364 ( 250 113 ) 2.5, 6 5.22 ( 0.71 ) 5.22 ( 0.71 ) 5.24 ( 0.72 ) -.01 Jail officers should encourage inmates to report consensual sexual acts that occur in jail b 366 (251, 114) 1, 6 4.60 (1.39) 4.59 (1.40) 4.60 (1.41) .02 Jail officers should encourage in mates to report sexual assaults that occur in jail b 369 (254, 114) 2, 6 5.48 (0.71) 5.47 (0.70) 5.49 (0.74) -.01 Jail officers should do everything they can to prevent consensual sexual acts in jail b 368 (254, 113) 1, 6 5.22 (0.99) 5.22 (0.99) 5.21 (1.00) .01 Jail officers should do everything they can to prevent sexual assaults in jail b 368 (253, 114) 2, 6 5.60 (0.62) 5.57 (0.67) 5.65 (0.50) -.07 110 Willing to respond: talk to inmates scale (Respond 2) a 368 ( 253 114 ) 1, 6 4.29 ( 1.33 ) 4.29 ( 1.35 ) 4.30 ( 1.31 ) -.02 Jail officers should talk to inmates about consensual sexual acts to discourage those activities b 368 (253, 114) 1, 6 4.18 (1.49) 4.18 (1.48) 4.17 (1.50) .01 Jail officers should talk to inmates about the risk of sexual assault in jail b 368 (253, 114) 1, 6 4.41 (1.42) 4.39 (1.42) 4.44 (1.43) -.05 Willing to respond: proactive action scale (Respond 3) a 363 ( 249 113 ) 1, 5 4.83 ( 0.97 ) 4.81 ( 1.02 ) 4.87 ( 0.85 ) -.08 Jail officers should use cell assignments to safe-guard inmates from sexual assault b 365 (251, 113) 1, 6 4.92 (1.04) 4.88 (1.07) 4.98 (0.96) -.10 Jail officers should refer inmate s to protective custody to safeguard them from sexual assault b 366 (251, 114) 1, 6 4.72 (1.20) 4.71 (1.25) 4.75 (1.09) -.05 a Codes: 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongl y disagree, b Reverse-coded item, c Codes: 1 = always, 2 = generally, 3 = sometimes, 4 = rarely, 5 = never, d Mean difference tested through independent samples t -test, p <.05, ** p <.01

PAGE 111

Table 4-4. Descriptive statistics by education: dependent variables N = 376 (none, degree) Min, max a Full sample mean (SD) No degree n = 219 mean ( SD ) Degree n = 156 mean (SD) Mean difference d Victim blaming a 374 ( 218, 155 ) 1, 6 2.07 ( 0.97 ) 2.02 ( 0.94 ) 2.14 ( 1.01 ) -.10 Inmates who have previously con sented to participate in sexual acts in jail get what they deserve if they are raped by other inmates b 375 (219, 155) 1, 6 2.34 (1.34) 2.29 (1.29) 2.41 (1.41) -.13 Some inmates deserve to be raped in jail because of the way they act b 375 (219, 155) 1, 6 1.96 (1.07) 1.91 (1.03) 2.03 (1.12) -.11 Homosexual inmates get what they deserve if they are raped in jail b 375 (219, 155) 1, 6 1.90 (1.02) 1.85 (0.97) 1.97 (1.10) -.11 Inmates who take money or cigarettes in exchange for consensual sexual acts get what they deserve if they are raped by other inmates b 374 (218, 155) 1, 6 2.21 (1.26) 2.14 (1.23) 2.31 (1.31) -.16 111 Inmates who dress or talk in feminine ways get what they deserve if they are raped in jail b 374 (218, 155) 1, 6 2.01 (1.14) 1.98 (1.10) 2.06 (1.19) -.07 Most men who are raped by a man are somewhat to blame for not being more careful b 374 (218, 155) 1, 6 2.01 (1.08) 1.97 (1.05) 2.06 (1.14) -.08 Most men who are raped by a man are to blame for not escaping or fighting off the man b 374 (218, 155) 1, 6 2.08 (1.12) 2.05 (1.11) 2.12 (1.15) -.06 Credibility of victims c An inmate tells you he was raped. How likely are you to believe him if he: 358 (210, 147) 1.44, 5 3.55 (0.71) 3.56 (0.68) 3.54 (0.74) .02 Is a muscular inmate b 364 (212, 151) 1, 5 3.34 (1.05) 3.35 (1.02) 3.31 (1.09) .02 Is a homosexual inmate b 364 (212, 151) 1, 5 3.77 (0.89) 3.73 (0.90) 3.81 (0.87) -.08 Is a young inmate b 364 (212, 151) 2, 5 4.08 (0.65) 4.09 (0.65) 4.07 (0.67) .03 Is a gang member b 364 (212, 151) 1, 5 3.34 (1.06) 3.36 (1.04) 3.31 (1.09) .05

PAGE 112

Table 4-4. Continue d N = 376 (none, degree) Min, max a Full Sample Mean (SD) No degree n = 219 Mean ( SD ) Degree n = 156 Mean (SD) Mean differenced Is an inmate who owes money b 364 ( 212 151 ) 1, 5 3.76 ( 0.84 ) 3.80 ( 0.80 ) 3.72 ( 0.88 ) .08 Is drunk or high on drugs b 359 ( 210 148 ) 1, 5 3.29 ( 0.99 ) 3.29 ( 0.99 ) 3.29 ( 1.00 ) .01 Delayed reporting the incident b 362 ( 210 151 ) 1, 5 3.53 ( 0.90 ) 3.53 ( 0.86 ) 3.53 ( 0.95 ) -.00 Has previously consented to sex with other inmates b 361 ( 210 150 ) 1, 5 3.30 ( 1.01 ) 3.31 ( 1.00 ) 3.27 ( 1.04 ) .05 Has mental health problems b 362 ( 210 151 ) 1, 5 3.48 ( 0.88 ) 3.51 ( 0.86 ) 3.46 ( 0.91 ) .05 Rape definitions a 359 ( 207 151 ) 1, 6 4.81 ( 1.01 ) 4.79 ( 0.95 ) 4.84 ( 1.09 ) -.07 112 Inmate Jones physically overpowers inmate Smith. Smith has sex with Jones. Smith has been raped. b 363 (211, 151) 1, 6 5.16 (0.98) 5.11 (1.02) 5.21 (0.94) -.10 Inmate Jones threatens to tell other inmates that inmate Smith is an informant unless inmate Smith engages in sexual acts. Smith has sex with Jones. Smith has been raped. b 366 (213, 152) 1, 6 4.74 (1.33) 4.70 (1.29) 4.78 (1.40) -.09 Inmate Jones tells inmate Smith he will kill him unless Smith has sex with Jones. Smith has sex with Jones. Smith has been raped. b 367 (214, 152) 1, 6 5.08 (1.09) 5.10 (1.06) 5.05 (1.14) .05 Inmate Smith is an informant. Inmates Jones provides protection for Smith but demands that Smith participate in sexual acts. Smith has sex with Jones. Smith has been raped. b 363 (210, 152) 1, 6 4.49 (1.40) 4.43 (1.38) 4.58 (1.44) -.15 Inmate Jones loans inmate Smith money or some goods. Smith cannot pay Jones back. Jones tells Smith that he can participate in sexual acts to pay off his debt or inmate Jones will beat inmate Smith severely. Smith has sex with Jones. Smith has been raped. b 364 (211, 152) 1, 6 4.58 (1.40) 4.58 (1.35) 4.59 (1.46) -.02

PAGE 113

113 Table 4-4. Continue d N = 376 (none, degree) Min, max a Full sample mean (SD) No degree n = 219 mean ( SD ) Degree n = 156 mean (SD) Mean differenced Willing to respond: report and prevent scale (Respond 1) a 364 ( 212 151 ) 2.5, 6 5.22 ( 0.71 ) 5.29 ( 0.65 ) 5.13 ( 0.79 ) .14 Jail officers should encourage inmates to report consensual sexual acts that occur in jail, b 366 (213, 152) 1, 6 4.60 (1.39) 4.69 (1.31) 4.45 (1.50) .21 Jail officers should encourage in mates to report sexual assaults that occur in jail, b 369 (215, 153) 2, 6 5.48 (0.71) 5.49 (0.69) 5.46 (0.74) .04 Jail officers should do everything they can to prevent consensual sexual acts in jail b 368 (214, 153) 1, 6 5.22 (0.99) 5.33 (0.85) 5.07 (1.15) .25* Jail officers should do everything they can to prevent sexual assaults in jail b 368 (215, 152) 2, 6 5.60 (0.62) 5.62 (0.59) 5.57 (0.66) .05 Willing to respond: talk to inmates scale (Respond 2) a 368 ( 215 152 ) 1, 6 4.29 ( 1.33 ) 4.23 ( 1.35 ) 4.38 ( 1.31 ) -.13 Jail officers should talk to inmates about consensual sexual acts to discourage those activities b 368 (215, 152) 1, 6 4.18 (1.49) 4.14 (1.51) 4.22 (1.46) -.06 Jail officers should talk to inmat es about the risk of sexual assault in jail b 368 (215, 152) 1, 6 4.41 (1.42) 4.32 (1.42) 4.53 (1.42) -.20 Willing to respond: proactive action scale (Respond 3) a 363 ( 213 149 ) 1, 5 4.83 ( 0.97 ) 4.71 ( 1.01 ) 4.99 ( 0.89 ) -.25* Jail officers should use cell assignments to safe-guard inmates from sexual assault b 365 (215, 149) 1, 6 4.92 (1.04) 4.80 (1.11) 5.09 (0.91) -.28** Jail officers should refer inmate s to protective custody to safeguard them from sexual assault b 366 (213, 152) 1, 6 4.72 (1.20) 4.62 (1.23) 4.86 (1.15) -.23 a Codes: 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongl y disagree, b Reverse-coded item, c Codes: 1 = always, 2 = generally, 3 = sometimes, 4 = rarely, 5 = never, d Mean difference tested through independent samples t -test, p <.05, ** p <.01

PAGE 114

CHAPTER 5 PERCEPTIONS OF VICTIMS QUANTITATIVE RESULTS Introduction This chapter explores two of the four resear ch questions regarding correctional officers attitudes about sexual assault and sexual assault victims outlined in chapter three: correctional officer attitudes toward victim blaming and inma te credibility. For each of these dependent variables, the results are presented using at least three separate analyses. First, paired sample t tests examine whether there are significant diffe rences across the sample means for various questions in the blaming and credibility indices. Second, a correlation analysis examines whether certain independent variables are significan tly associated with each dependent variable. The third analysis is a theoretically built, forc ed-entry stepwise ordinary-least squares (OLS) regression model using several st eps to examine change in variance and coefficients across model. The final analysis, where applicable, pres ents separate regression models by sex and uses coefficient comparison tests ( z -tests) to examine the effects of significant indepe ndent variables across sex. Victim Blaming Recall from Chapter 3 that the first resear ch question is whether correctional officers attribute blame to victims of rape in jail. First, the results of a paired samples t -test compare relative levels of blaming across characteristics of inmates. Second, each independent variable is entered into a correlation analysis with victim blaming in order to determine which independent variables are significantly associated with blaming attitudes. Finally, th e independent variables that are significantly correlated with victim blam ing are entered in a theoretically driven forcedentry stepwise regression model. 114

PAGE 115

Blaming and Inmate Characteristics This section reports the findings of a paired samples t -test to examine the mean differences in blaming specific types of inmates among the fu ll sample. As previously discussed, a paired t test is used as each respondent in the sample was asked to indicate his or her response for each blaming item; therefore responses for one item may be dependent on responses for other items (Agresti & Finlay, 1997). I compared the sample means for each victim blaming item with each other victim blaming item. It is apparent from the paired samples t -test that in comparing mean levels of blame, inmates who have previously consented to sex with other inmates are seen as the most blameworthy (Table 5-1). Recall from Chapte r 4 that higher response values for the victim blaming index and individual items indicate an increas ed level of blame attributed to inmates. In other words, a response of six indicates that the respondent strongly agrees that the inmate is to blame for rape, while a response of one indicates that the respondent str ongly disagrees that the inmate is to blame. Officers attribute significantly higher levels of blame to these inmates compared to inmates who are homosexual, take goods (money or cigarettes) in exchange for sexual acts, dress or talk in feminine ways, ar e not more careful, do not escape or fight, and who act a certain way. Officers in this sample also perceive inma tes who take goods in exchange for sexual acts as more blameworthy than other inmates (Table 51). They report significantly greater levels of blaming for inmates who take money or cigarettes in exchange for sexual acts than for inmates who are homosexual, feminine, are not more careful, do not escape or fi ght, and act a certain way. Inmates who do not escape or fight are seen as the next most blameworthy. Officers report significantly higher means for these inmates than they did for inmates who are homosexual or inmates who act a certain way. The sample re ports significantly higher means for blaming inmates who are not more careful than for blam ing homosexual inmates. Inmates who dress or 115

PAGE 116

talk in feminine ways are only seen as significantly more blameworthy than homosexual inmates are. Where significant differences exist, inmates w ho act a certain way (deserve to be raped because of the way they act) are never seen as more blameworthy compared to other inmates (Table 5-1). One possibility for this finding is that the question does not refer to any specific trait or behavior of an inmate and is somewhat vague. Finally, officers rate homosexual inmates with significantly lower means (less blame) than all other types of inmates. Specifically, homosexual inmates are regarded as less blam eworthy than inmates who have previously consented to sexual acts, take money or cigarettes in exchange for sexual acts, dress or talk in feminine ways, are not more careful, do not escape or fight, and who act a certain way. These findings imply that officers in this sample see so me inmates as more blameworthy than others. Specifically, inmates who have previously consen ted to sexual acts or who take money or cigarettes in exchange for sexua l acts are attributed the most bl ame. Inmates who do not escape or fight are also seen as rela tively blameworthy. Inmates who are not more careful or who dress or talk in a feminine manner ar e only considered more blamewor thy than homosexual inmates. Officers in this sample were least likely to attr ibute blame to inmates who are homosexual. This is an interesting finding, and is di scussed in depth in Chapter 8. Bivariate Correlations The victim blaming zero-order correlations included in Tabl e 5-2 are based on what prior studies have indicated may be important in pr edicting attitudes toward inmates generally, or blaming attitudes specifically. Some studies of correctional officers imply that seniority, job satisfaction, and levels of stress (Arthur, 1994; Farkas, 1999; Jurik, 1985; Poole & Regoli, 1980a; Teske & Williamson, 1979) affect correctional offi cer attitudes, none of these variables is significantly correlated with blaming inmate victims of rape. In addition, facility size (inmate 116

PAGE 117

count) is included in the bivari ate correlations and is signif icantly associated with victim blaming. Although some prior research indicates that age, education, religiosity, and race might influence attitudes toward rape victims (Bur t, 1980; Jimenez & Abreu, 2003; Lee & Cheung, 1991; Mori, Bernat, Glenn, Selle, & Zarate, 1995 ; Nagel, Matsuo, McIn tyre, & Morrison, 2005; Sheldon & Parent, 2002; Xenos & Smith, 2001), these demographic variables are not bivariately associated with blaming among this sample. Sex is incorporated in the analysis as most studies that have examined the effects of respondent sex have found that males have a greater tendency to victim blame than females do (Barnett et al ., 1992; Calhoun et al., 1976; Cann et al., 1979; Feild, 1978; Fulero & Delara, 1976; Feldman-Summ ers, & Linder, 1976; Kanekar & Kolswalla, 1980; Kleinke & Meyer, 1990; MaCrae & Shephe rd, 1989; Pollard, 1992; Thornton et al., 1981; Thornton et al., 1982; Thornton & Ryckman, 1986; Whatley, 2005). Attitudes toward homosexuality is included in th e analysis as research on non-incarcerated sexual assault victims indicates that homosexual victims are often seen as more culpable, especially male homosexual victims (Davies et al., 2006; Davies & Rogers, 2006; Ford et al., 1998; Mitchell et al., 1999). Other research ha s found greater tendencies to blame among those with less favorable attitudes toward homo sexuality (Anderson, 2004; Burt & DeMello, 2002; Wakelin & Long, 2003). In addition, one study that examined victim blaming as a dependent variable in an analysis of prison correctional officer per ceptions found that opposition to homosexuality increases th e probability of victim blaming (Eigenberg, 2000a). Although perception of inmate cred ibility is incorporated as a dependent variable later in the analysis, it is also included as an independent variable in the analysis of victim blaming attitudes. Research suggests that personal traits and behaviors of victims can influence 117

PAGE 118

credibility or believability of reports of sexual assault (Feldman-Summers & Linder, 1976; Jordan, 2004; Kanekar & Kolsawalla, 1980; Krulewitz & Nash, 1980; McCaul et al., 1990; Pollard, 1992; Pugh, 1983; Spears & Spohn, 1996), a nd the current study incorporates victim traits into the measure of in mate credibility. Because res earch suggests that personal characteristics of female victims of rape and sexual assault influence blaming attitudes (Bell et al., 1994; Bridges, 1991; Pugh, 1983; Schuller & Has tings, 2002), it is expected that whether respondents perceive inmates as credible may also influence whether they perceive a victim as responsible. No prior research examines how professional orientation aff ects the blaming attitudes of officers in jails, though one study suggests that victim blaming by officers is prisons is correlated with less preference for social distance from inmates and punitiveness (Eigenberg, 2000a). Though little prior research on at titudes toward victims has in cluded professional orientation items as independent variables, these indices (c ounseling roles, punitive orientation, corruption of authority, and social distance) are incorporated here as measur es of officer perceptions about their profession and inmates as one study indicates that professional orientation may influence officer definitions of rape (Eigenberg, 2000a). The analysis also incorporates attitudes about rape myth items, as female rape myths are infl uential in research about blaming female victims of rape (Frese et al., 2004; Jenkins & Dambrot, 1987; Kopper, 1996; Mason et al., 2004). To my knowledge, no studies of blaming inmate victims of sexual assault have incorporated male rape myths as a predictor. Initial zero-order correlations i ndicate that victim blaming attitudes are associated with sex, attitudes toward homosexuality, inmate credibilit y, counseling roles, punitive orientation, social distance, and rape myths. Specifically, higher levels of victim blam ing are significantly 118

PAGE 119

associated with being male (r = -.22, p<.01), negative attitudes toward homosexuality (r = -.27, p<.01), disbelief in inmate credibility (r = .24, p<.01), lack of support for counseling roles (r = .15, p<.01), punitive orientation (r = .24, p<.01), support for social distance from inmates (r = .13, p<.01), and male rape myth acceptance. Only two of the four male rape myths have significant associations with vi ctim blaming (it is impossible for a man to rape a man and most men who are raped do not ne ed counseling after th e incident) (Table 5-2). These two items have significant correlations with victim blaming (r = .32, p<.01 and r = .30, p<.01, respectively). Multivariate Analysis Only variables that are significantly correlated with victim bl aming at the bivariate level are entered into the stepwise model. The regression model for victim blaming consists of six steps(Table 5-3). Each step is displayed in order from top to bottom in the regression tables; moreover, the tables present all variables incl uded in each step of the analysis. The model incorporates sex in the first step and facility size (inmate count) in the second step. Attitudes toward homosexuality in the third step1. These variables are entered in the initial steps in order to control for demographics, facility characteri stics, and predictors found important in prior research, and to examine the relative c ontribution of subsequent variables. 1 A follow-up analysis is conducted by splitting the regression models by sex and using coefficient comparison tests ( z-tests) to determine statistically signi ficant differences (see Clogg, Petkova, & Haritou, 1995 and Paternoster, Brame, Mazerolle, & Piquero, 1998). Prior studies suggest that the tendency to victim blame varies by sex and the zero-order correlation results suggest that sex is significantly and moderately (r = .46, ** p <.01) correlated with attitudes towards homosexuality (Table 5-2). Also, both va riables are significantly associated with victim blaming and the t-test results discussed in chapter 4 indicate that victim blaming and attitudes towards homosexuality are significantly different across sex in this sample (Tables 4-1 and 4-3). Alth ough these results suggest that the strength of attitudes towards homosexuality as a predictor of victim blaming likely varies by sex, the z -tests reveal no significant difference betw een coefficients or in the effect of attitudes towards homosexuality on victim blaming across sex. 119

PAGE 120

The fourth step includes perceptions of inmate credibility; this variable is included next to gauge its relative importance while controlling for other variables as prior rape research suggests that perceptions of victims ma y influence victim blaming. The next step includes professional orientation items that were significant predicto rs of victim blaming at the bivariate level: counseling roles, punitive orientation, and soci al distance. Although onl y punitive orientations and social distance were associated with (but not predictors of) victim blaming in previous research examining prison samples (Eigenberg, 2 000a), all professional orientation indices were originally included in this analysis in order to assess their contribution in this jail sample. The final step includes the two male rape myth items th at are significant at the bivariate level. Male rape myth items, to my knowledge, have not been included in tests of victim blaming attitudes with correctional officer samples. These variable s are entered last in order to examine the unique contribution of items shown to affect attitudes toward female rape victims. Step one indicates that sex is a significant predictor of victim blaming, specifically, males are significantly more likely to victim blame than females (Table 5-3). In step two, facility size is significant and suggests that officers working in facilities with large inmate counts are more likely to victim blame. Step three indicates that, while controlling for sex and facility size, officers less tolerant of homosexua lity are significantly more likely to blame inmate victims of rape. Sex loses significance in this step; this result may indicate that variance in blaming attitudes may be better explained by beliefs about homosexuality than by sex. While sex explains 5% and facility size (inmate count) explains 2% of the variance in victim blaming, attitudes toward homosexuality explains an additional 8% of variance in the dependent variable. When perception of inmate credibility is incl uded in step four, facility size and attitudes toward homosexuality remain significant while sex remains insignificant. Th is step reveals that, 120

PAGE 121

controlling for other variables, t hose who attribute credibility to inmates who report rape are less likely to victim blame. Adding perceptions of inmate credibility to the model explains an additional 3% of variance in victim blaming. This result is not surprising as many of the credibility items (Table 3-5) in corporate traits of carelessness or irresponsibility, items that Pollard (1992) also found to increase blaming of female victims of rape. Step five incorporates the pr ofessional orientatio n items that are bivariately related to victim blaming. While facility size, attitude s toward homosexuality, and belief in inmate credibility remain significant in this step, the only professional orient ation item that is an important predictor of blaming is punitive orie ntation. Specificall y, those with punitive orientations are more likely to blame inmate vict ims of rape. The inclus ion of these variables adds an additional 6% of varian ce in the model predicting victim blaming. Finally, inclusion of the two male rape myth items explains an addi tional 11% of variance in victim blaming. Both items are significant predictors of victim blaming, specifically, t hose who adhere to rape myths that it is impossible for a man to rape a man a nd that raped men do not need counseling after the incident are more likely to victim blame. Examination of the standardized coefficients across steps in the model reveals that male rape myths, attitudes toward homosexuality pun itive orientation, and facility size (inmate count) are the strongest predictors of victim blaming among this sample. These items are also consistently significant across all step s in the model. Based on both the R2 change and standardized coefficients, male rape myths appear to be the st rongest predictors of victim blaming, followed by attitudes toward homose xuality and punitive orientation, respectively2. 2 Results were estimated using both listwise deletion an d expectation-maximization (EM) imputation methods to address missing values in incomplete cases. Although th e results did not change dramatically, the analysis with imputed data indicates sex as a signifi cant predictor of victim blaming. Becau se the true values of the missing data are unknown, however, the results presented in Table 5-3 use listwise deletion to address missing values. 121

PAGE 122

Though not the most important, perceptions of in mate credibility and f acility size are also important predictors in explaini ng victim blaming. The model explains 34% of the variance in victim blaming among this sample. Summary of Victim Blaming Analysis Respondents in this sample are similar to samp les elsewhere that assi gn relative levels of blame to different types of victims, especially inmate victims (Eige nberg, 1989). While sexual orientation may be a factor in blaming non-incarcerated victims of sexual assault in previous studies, it is not an important f actor here. What makes inmates significantly more culpable in this study is having previously consented to sex with other inmates or having exchanged sex for goods such as money or cigarettes. Respondents w ho work in larger facilities, are less accepting of homosexuality, do not see inmates who report rape as credible, have punitive orientations, and adhere to male rape myths are more likely to blame inmate victims of rape. The effect of attitudes toward homosexuality is not surprisi ng, as those with less accepting attitudes about homosexuality are more likely to blame in othe r studies as well. However, it appears that attitude toward homosexuality is a stronger predictor of victim blam ing attitudes than sex in this model. This is interesting, as pr evious research has consistently shown that sex is also a rather strong predictor of blaming attitudes. I will exam ine further the effect of sex on perceptions of inmate credibility as a dependent variable later in this analysis. Inmate Credibility The second research question explored in this chapter examines how officers perceive the credibility of victims of sexual assault in jail. Specifically, are all inmates equally credible as victims of sexual assault, or ar e there certain inmates that are more likely to be believed as victims? In order to explore thes e research questions, the first part of the analysis uses a paired samples t -test to examine whether there are signifi cant differences in perceptions of inmate 122

PAGE 123

credibility by type of inmate. The results of a zero-order correlation are presented to examine which variables are significantly associated with perceptions of inmate credibility. The analysis also examines the impact of several variables on perceptions of inmate credibility in a forcedentry, theoretically driven stepwise regression model. A follow-up analysis is conducted by splitti ng the regression models by sex and using coefficient comparison tests ( z -tests) to determine statistically significant differences (see Clogg, Petkova, & Haritou, 1995 and Paternoster, Brame, Mazerolle, & Piquero, 1998). This analysis is conducted as prior studies suggest that attitudes about victims vary by sex and the zero-order correlation results suggest that sex is significantly and moderately (r = .46, **p<.01) correlated with attitudes toward homosexuality (Table 5-2). In addition, both variables are significantly associated with perceptions of credibility and the t -test results discussed in Chapter 4 indicate that some credibility items and attitudes toward homosexuality are significantly different across sex in this sample (Tables 4-1 and 4-3). Although these results suggest that the strength of attitudes toward homosexuality as a predictor of perceptions of credibility likely varies by sex, the z -tests reveal no significant difference between coefficients or in the effect of attitudes toward homosexuality on perceptions of credibility across sex Credibility and Inmate Characteristics There are nine statements included in the survey that ask respondents to indicate how likely they are to believe certain inmates who re port rape (Table 4-3). I use paired sample t -tests to examine the mean differences between likelih ood of believing each type of inmate compared to each other type of inmate among the full sample. A paired sample t -test is appropriate as each respondent indicated their response for each item; therefore, responses for one item may be dependent on responses for othe r items (Agresti & Finlay, 1997). 123

PAGE 124

The results of the paired sample t -tests indicate that there are significant mean differences regarding credibility of many of the inmate types pr esented in the survey (Table 5-4). Recall that these items are coded so that hi gher values represent greater tende ncy to believe reports of rape from that victim. A response value of five indica tes that they would alwa ys believe a report of rape and a response value of one indicates that th ey would never believe a report of rape from that inmate. The mean for a muscular inmate is significantly lower than the mean for a homosexual inmate, for example, which signifies that the sample is significantly more likely to believe reports of rape from the homosexual in mate than the muscular inmate. In the comparisons with significant differences, the means for muscular inmates are always lower than the means for other inmates (homosexuals, the young, inmates who ow e money, inmates who delay reporting, and inmates with mental health problems). This indicates that, compared to other inmates, respondents are significantly less likely to believe muscular inmates who report being victims of rape. There are no significant differences between mean credibility of muscular inmates and gang members. In comparisons where there were significant differences between gang members and others (homosexuals, the young, inmates w ho owe money, inmates who delay reporting, and inmates with mental health problems), officers are also significantly less likely to believe gang members rape allegations. The perception of cr edibility is also significantly lower for drunk or high inmates than for other inmates (homose xuals, the young, inmates who owe money, inmates who delay reporting, and inmates with mental health problems). There are no significant differences between the mean credibility for drun k or high inmates and mu scular inmates or gang members. 124

PAGE 125

Young inmates, however, are signif icantly more likely to be believed as victims of rape than every other inmate (Table 5-4). In most comparisons with significant differences, homosexual inmates are also more likely to be believed than other inmates. Compared to young inmates, however, homosexual inmate s are significantly less likely to be perceived as credible. Other inmates perceived as credible in most comparisons are those who owe money. Inmates who owe money are significantly more likely to be believed in comparisons with muscular inmates, gang members, drunk or high inmates, those who delay reporting, inmates who have previously consented to sex with other inmates, and inmates with mental health problems. There are no significant differences between inma tes who owe money and homosexual inmates. Inmates who delay reporting a ra pe incident and inmates with mental health problems are perceived as similarly credible. These inmates have significantly higher means and are more likely to be believed than muscular inmates, gang members, inmates who were drunk or high, and inmates who had previously consented to sex with other inmates. Inmates who delay reporting and those with mental health problems are not, however, more likely to be believed than homosexuals, young inmates, or inmates who owe money. Finally, inmates who have previously consented to sex with other inmates ar e significantly less likely to be believed in comparisons with homosexuals, young inmates, inmates who owe money, and inmates who delay reporting. These results imply that inmates that may be perceived as strong or powerful in a jail context (muscular inmates, gang members) are fa r less credible as rape victims than other inmates. Drunk or high inmates or those who have previously consented are also perceived as consistently less credible than other victims, and are not perceived as differentially credible compared to muscular inmates or gang member s. Respondents consistently perceive young 125

PAGE 126

inmates and homosexuals as credible. Also rega rded as credible were inmates who owe money. This is an interesting finding, and possibly suggests that inmate s who owe money are targeted for victimization. Respondents are less consistent regarding the cr edibility of inmates who delay reporting or those with mental health problems, however. Though perceived as more credible than inmates possibly perceived as either strong (muscular, gang members) or unlikely (drunk, previously consented), they are less credible than those who might be perceived as weaker or obvious targets (the young, homosexuals, inmates who owe money). These results imply that respondents rank the credibility of inmates who report rape in a rela tively consistent order, with young inmates as the most credible, followed by homosexual inmates and inmates who owe money. Inmates who delayed reporting and inma tes with mental health problems are more credible than some inmates, yet less credible th an others. Finally, inmate s perceived as the least credible are muscular inmates, gang members, inmates who were drunk or high on drugs and inmates who had previously consented to sex with other inmates. These findings are consistent with those from other studies of both in carcerated and non-incarcerated victims. Bivariate Correlations The variables entered in the zero -order correlation analysis are the same variables used to examine the victim blaming dependent variable. The analysis was conducted in this way as most research on credibility of victims examines how legal or contextual factors of the incident influence respondent perceptions of credibility. In other words, ther e is little research about what characteristics of respondents might influence per ceptions of credibility re garding rape victims. Though attitudes about victim blaming and perceptions of credibility may be explained by different factors, the same variables are examined here as both dependent variables deal with perceptions about inmate victims. 126

PAGE 127

The research that does examine characteris tics of respondents indicates that female subjects may be more prone to believe victim accounts of sexual assau lt (Bottoms & Goodman, 1994; Gabora, Spanos, & Joab, 1993; Goodma n et al., 1989; Jackson & Nuttall, 1994; ODonohue et al., 1992). Another f actor that may influence percep tions of credibility discussed in Chapter 2 is occupational or educational status, (ODonohue et al., 1998). In this study, however, education is not signifi cantly correlated with perceptions of credibility. In addition, age, religiosity, and race are not bivariately associated with perceptions of inmate credibility. The variables that are significantly associated with perceptions of inmate credibility are sex (r = .19, p<.01), stress (r = .11, p<.05), attitudes toward homosexuality (r = .18, p<.01), victim blaming (r = -.26, p<.01), social distance (r = -.14, p<.05) (Table 5-5). In addition, two rape myth items are also significantly correlated with credibility: it is impossible for a man to rape a man (r = -.16, p<.01) and men who are raped by a ma n do not need counseling after the incident (r = -.12, p<.05). Recall that these two male ra pe myths were also associated with victim blaming. None of the remaining professional orientation or male rape myth items are significantly associated with perc eptions of inmate credibility. Multivariate Analysis There are six steps entered in the stepwise regression model for pe rceptions of inmate credibility (Table 5-6). Similar to the victim blaming model, sex is entered first in the full model. Because the model is analyzed separately by sex, the primary step in the models split by sex incorporates step two of the full model. The second step for the full sample (step one of the models split by sex) incorporates stress and attitudes toward ho mosexuality is entered in step three of the full model and the step two of the m odels split by sex. Step four of the full model (the third step in the models split by sex) inco rporates victim blaming. Step five of the full model (the fourth step in the separate male and female models ) includes social distance. The 127

PAGE 128

final step of all models adds the two male rape myths to determine their relative significance, controlling for other variables. The variables were entered in this way to control for factors that are significant in perceptions of victim culpability and blaming (sex and attitudes toward homosexuality) while examining the unique contributions of other factors. Full sample In the first step, sex is significant and e xplains 4% of the variance in the dependent variable. Females are more likely to perceive inmates who report ra pe as credible (Table 5-6). Step two indicates that stress is also significant, but only adds 1% of explained variance to the model. Officers who indicate greater amounts of job stress are more likely to perceive inmate reports of rape as credible. In step three, attitudes toward homosexuality does not add much variance to the model (1%), and is not significan t (though it was bivariately significant, it loses significance in the multivariate model). Sex and stress remain significant in step four, and the inclusion of victim blaming explai ns an additional 4% of variance in perceptions of credibility. Victim blaming is significant, and respondents w ho are less likely to victim blame are more likely to perceive inmates who report rape as credible (see victim blaming results presented earlier in this chapter). In step five, social distance adds 1% of explained variance an d is significant, while sex, stress, and victim blaming maintain their significance. Those who have less preference for social distance in working with inmates are more likely to perceive inmate reports of rape as credible. In step six, the inclusion of male rape myths expl ains less than1% of additional variance in the dependent variable. Although sex, stress, vic tim blaming, and social distance maintain significance in the final equation, neither male rape myth is significant here. The full model explains 11% of variance in perceptions of inma te credibility among this sample. Across the full sample equations, victim blaming, sex, stress, an d social distance appear to be consistent 128

PAGE 129

predictors of perceptions of inma te credibility. Examination of R2 change and standardized coefficients suggests that victim blaming is th e strongest predictor of perceptions of inmate credibility3, followed by sex, stress, and social distance. It is apparent th roughout the model that respondents who are less likely to victim blame, females, those who experience job stress, and those less likely to prefer social distance from in mates are more likely to perceive inmates who report rape as credible. Split samples The models are analyzed separately by sex b ecause the results from Chapter 4 indicate that males and females are significantly different in their perceptions of inma te credibility (for both the credibility index and five credibility items) In addition, both se x and attitudes toward homosexuality are significantly corr elated with perceptions of in mate credibility, and they are significantly associated with each other (Table 5-4). Male sample. The next part of the analysis separate d the sample by sex, and therefore the stepwise models for both males and females inco rporate only four step s (step one, previously incorporating the variable sex, is eliminated). Similar to the full sample, stress is also significant in the model with male respondents, and explains about 2% of the variance in the dependent variable. Attitudes toward homosexuality explains 3% of the variance in perceptions of inmate credibility and is significant (Table 5-6). Unlike the full sample, those who are more likely to accept homosexual lifestyles are significantly more like ly to believe inmate reports of rape. This 3 Although the full model results with expectation-maximization and regression imputation are somewhat different than those found using listwise deletion regarding relative strength of coefficients, the results are substantively similar regarding explained variance, R2 change and the strongest predictor, vi ctim blaming. The effects of social distance and sex are insignificant using listwise deletion, however. Whether this result is a function of listwise deletions tendency to bias estimates and produce Type II error is unknown, as the values of the missing data are unknown. Though this finding may indicate that cases w ith missing data are substantively different from complete cases, the results for the male and female models are more similar across method, and may be more reliable when analyzed separately (see below). 129

PAGE 130

variable maintains significance in the next ste p, and the inclusion of victim blaming in the equation improves explained variance by 2 percent. Similar to the full sample, victim blaming is significant and indicates that those who are more likely to blame victims are less likely to perceive inmate reports of rape as credible (see victim blaming re sults presented earlier in this chapter). Adding social distance to the next equation (step five for the full sample) only improves explained variance by 1%, and unlike the full sample results, this variable is not a significant predictor of perceptions of inmate credibility for male respondents. In the final step, male rape myths explain an additional 3% of variance in perceptions of inmate cr edibility, though only one male rape myth is significant. Male respondents who adhere to the male rape myth that it is impossible for a man to rape a man are signific antly less likely to perceive inmates who report rape as credible. Inclusion of male rape myths in the model makes victim blaming insignificant, suggesting that beliefs about male rape myths may be a more powerful predictor of whether officers perceive inmate reports of rape as credible. Recall that attitudes toward homosexuality is consistently insignificant across the full sample equations. Likewise, where victim blami ng and social distance are significant in the full model, they are not significant pr edictors in the final model for male respondents. Victim blaming may be a strong and significant predic tor in early equations for male respondents, however, in the final equation it appears that varian ce in perceptions of inmate credibility may be better explained by adherence to rape myths, a ttitudes toward homosexuality, and stress than by tendencies to victim blame for ma le respondents. Overall, this model explains 10% of the total variance in perceptions of inmate credibility among male respondents. 130

PAGE 131

Examination of the relative importance of va riables across equations indicates that one male rape myth item (it is impossible for a man to rape a man) is the strongest predictor of believing inmates who report rape for male respondents. Males who accept this myth are less likely to believe inmate reports of rape. Th is variable has relatively large standardized coefficients across equations and the largest change in R2, compared to the other variables in the model. Also important in predicting perceptions of inmate credibility among males is attitudes toward homosexuality and stress4. The results for the full model, however, indicate that victim blaming and social distance may play a more substa ntial role in predicting perceptions of inmate credibility for the full sample. Female sample. The first step of the female equation (step two for the full sample) indicates that, unlike the male sample, stress is no t a significant predictor of perception of inmate credibility for females and it explains little variance (less than 1%). In the second step of the female model (step three for the full sample), attitudes toward homosexua lity is not significant and explains less than 1% of the total variance in perceptions of inmate credibility (Table 5-6). In the third step (step four for the full sample), victim blaming explains an additional 16% of variance and is significant. Females who blame victims are significantly less likely to perceive inmates who report rape as credible. Victim blam ing remains significant in all other steps in the female model. In the fourth step (step five fo r the full sample), social distance does not explain much more variance (2%) in perceptions of credibil ity and is not significant. In the final step (step six for the full sample), only one male rape myth variable is significant. Male rape myths add 5% of explained variance to the model; females who adhere to the myth that raped men do not need counseling are less likely to perceive inmates as credible. 4 This is a result consistent with both listwise deletion results and regression imputation results. In addition, the results for the female model are also robust across these methods. 131

PAGE 132

Victim blaming adds the most predictive power to the model for female respondents. This item has the largest R2 change and standardized coefficients across equations. According to the R2 change and standardized coeffi cients, the male rape myth item is the next strongest predictor in the final model for female respondents. Acro ss the female sample equations, only one rape myth item (men who have been raped by a man do not need counseling afte r the incident) and victim blaming are significant predictors of pe rceptions of inmate credibility. Females who blame victims and who believe that men do not need counseling afte r the incident are less likely to perceive inmates who report rape as credible. Overall, this m odels explains more variance in perceptions of credibility for female respondents that for male respondents. A total of 23% of the variance in perceptions of inmate credibility among female respondents is explained by this model. Comparing model results across the full sample and the male and female sub-samples indicates that victim blaming is an important pr edictor of perceptions of inmate credibility for females and the full sample. Beliefs about male rape myths also matter for both females and males; however, these samples differ regarding the type of myth that is important. The significant effect of job stress on perceptions of inmate credibility appears relevant mostly for males and the full sample. Summary of Inmate Credibility Analysis The results of the inmate credibility analysis imply that respondents in this sample do rank some inmates as more credible when they report rape victimization than ot hers. In addition, they are consistent in ranking inmate believability. Young inmates are perceived as the most credible, followed by homosexual inmates and inmates who owe money. Those who delay reporting and have mental health problems are viewed as more credible than some inmates, yet less credible than others. The inmates perceived as the leas t credible are muscular, gang members, drunk or 132

PAGE 133

high inmates, and inmates who had previously consented to sex with other inmates. As discussed in detail in Chapter 8, these findings are consistent w ith those from other studies. Comparing results across multivariate models, it is apparent that victim blaming is a strong predictor of perceptions of inmate credibi lity for the full sample and female respondents. It is plausible that the results for the female respondents regarding the relative importance of victim blaming may be driving the results fo r the full model; victim blaming, though important in some steps, does not appear to be as impor tant as adherence to rape myths in explaining perceptions of inmate credibility for males. It is interesting that adherence to rape myths matters for both males and females; however, a different male rape myth predicts likelihood of believing inmate accounts of rape for each sex. Perceptions of inmate credibility are less likely for males who adhere to the myth that it is impossible for a man to rape a man and for females who adhere to the myth that males who are raped by other males do not need counseling after the incident. The implications of these finding are discussed in-depth in Chapter 8; however, these results support other research that finds that perceptions of rape victims may be partially influenced by sex and subs cription to rape myths. Conclusion The findings in this chapter s uggest that officers in this sa mple, like respondents in other settings, rank some victims as more culpable an d credible than others. Specifically, inmates who have previously consented in sex with inmates or who have taken items such as cigarettes or money in exchange for sex with other inmates ar e viewed as significantly more blameworthy in sexual assault incidents than othe r inmates. Other factors, such as not escaping or fighting or being feminine are not as important in attributing blame. Least important to this sample when it comes to assigning blame is sexual orientation. Ho mosexual inmates are not seen as particularly 133

PAGE 134

culpable for being sexually assaulted, compared to other inmates. In predicting blaming attitudes, respondents who were accepting of hom osexuality, who view inmates who report rape as credible, who are less punitive, and who do not subscribe to rape myths are less likely to blame inmate victims of sexual assault. Regarding credibility of rape victimizati on, inmates who are muscular inmates, gang members, or who have previously consented to se x with inmates are perceived as less credible than homosexuals or young. Similar to results with credibility of female vi ctims in other studies, inmates who are high or drunk are viewed as less credible when they report victimization. Respondents consistently perceive young inmates and homosexuals as credible; in addition, inmates who owe money are regarded with some degree of credibility. These results indicate that officers in this sample perceive the young and homosexuals as credible and the blaming results indicate they do not tend to blame homosexuals for sexual a ssault. This could imply that officers may expect young and homosexual inmate s to be targeted victims who should be believed when they report rape. In addition, inma tes who have previously consented to sex with other inmates are seen as more culpable and less credible than other inmates. The multivariate analysis reveals that in the full and female sa mples, a tendency to victim blame is a strong predictor of perceiving inmate repor ts of rape as credible. Adhere nce to the rape myth that men who are raped do not need counsel ing after the incident is also a strong predictor of blaming attitudes for females in this sample. For males in the credibility analysis perceptions of inmate credibility are less likely for thos e who adhere to the myth that i t is impossible for a man to rape a man and those who do not agree with homosexual lifestyles. These results are consistent w ith prior studies as percepti ons about homosexuality, rape myths, and victims drive beliefs about culpabil ity and credibility rega rding victimization. 134

PAGE 135

Female respondents perceive inmate victims as less blameworthy and more credible, similar to studies of non-incarcerated vict ims. In perceptions of male rape, negative attitudes toward homosexuality significantly predicts a tendency to victim blam e, however, regarding credibility, these attitudes are only relevant for male responde nts. In other words, males who hold negative attitudes toward homosexuality are less likely to perceive ma le inmates who report rape as credible. Whether respondents ac cept rape myths is a consistent determinant of blaming and perceptions of credibility. If respondents adhere to male rape myths, they are far more likely to blame inmate victims of rape and less likely to believe reports from these victims. How respondents in this study define rape in a corr ectional setting and whethe r they are willing to respond to these incidents will be explored furt her in Chapter 6. Following that, Chapter 7 will examine officer professional orientation and perceptions of inmate attitudes, and Chapter 8 will provide a discussion of all results. 135

PAGE 136

Table 5-1. Victim blaming paired sample t -tests Mean a (SD) Lower Upper t-value b Previously consented to participate in sexBecause of the way they act 2.34 (1.33) 1.95 (1.07) .29 .48 8.01** Previously consented to participate in sexHomosexual inmates 2.34 (1.33) 1.90 (1.02) .34 .54 8.72** Previously consented to participate in sexTake money or cigarettes in exchange for sex 2.34 (1.33) 2.21 (1.26) .03 .22 2.61** Previously consented to participate in sexDress or talk in feminine ways 2.34 (1.33) 2.02 (1.14) .22 .42 6.19** Previously consented to participate in sexNot being more careful 2.34 (1.33) 2.01 (1.08) .20 .45 5.16** Previously consented to participate in sexNot escaping or fighting 2.34 (1.33) 2.07 (1.12) .14 .39 4.08** Because of the way they act Homosexual inmates 1.95 (1.07) 1.90 (1.02) -.06 .13 1.32 Because of the way they act Take money or cigarettes in exchange for sex 1.95 (1.07) 2.21 (1.26) -.35 -.16 -5.42** Because of the way they act Feminine 1.95 (1.07) 2.02 (1.14) -.15 .03 -1.32 Because of the way they act Not being more careful 1.95 (1.07) 2.01 (1.08) -.16 .05 -1.09 Because of the way they act Not escaping or fighting 1.95 (1.07) 2.07 (1.12) -.23 -.01 -2.08* Homosexual inmates Take money or cigarettes in exchange for sex 1.90 (1.02) 2.21 (1.26) -.39 -.23 -7.60** Homosexual inmates Dress or talk in feminine ways 1.90 (1.02) 2.02 (1.14) -.18 -.05 -3.48** Homosexual inmates Not being more careful 1.90 (1.02) 2.01 (1.08) -.20 -.02 -2.45* Homosexual inmates Not escaping or fighting 1.90 (1.02) 2.07 (1.12) -.26 -.08 -3.66** Take money or cigarettes in exchange for sexDress or talk in feminine ways 2.21 (1.26) 2.02 (1.14) .12 .26 5.43** Take money or cigarettes in exchange for sexNot being more careful 2.21 (1.26) 2.01 (1.08) .09 .31 3.50** Take money or cigarettes in exchange for sexNot escaping or fighting 2.21 (1.26) 2.07 (1.12) .03 .25 2.44* Dress or talk in feminine waysNot careful 2.02 (1.14) 2.01 (1.08) -.09 .10 0.07 136

PAGE 137

137 Table 5-1. Continued Mean a (SD) Lower Upper t-value b Dress or talk in feminine waysNot escaping or fighting 2.02 (1.14) 2.07 (1.12) -.15 .04 -1.14 Not being more careful Not escaping or fighting 2.01 (1.08) 2.07 (1.12) -.12 .00 -1.82 a Reverse-coded item, 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongly disagree, b Mean difference tested through paired t -test, p <.05, **p <.01

PAGE 138

Table 5-2. Victim blami ng zero-order correlations Education Sex Age Race Religiosity Seniority Satisfaction Stress Count ATH a Victim blaming Education Sex -.01 Age .16** -.10 Race .06 .11* -.04 Religiosity -.03 .06 .18** .25** Seniority .16** -.11* .52** .06 .15** Satisfaction -.04 -.01 .10* -.03 .19** .00 Stress .02 .07 .13* .00 .08 .15** -.15** Count .11* .16** .05 -.06 -.07 -.03 .05 -.13* ATH .04 .46** -.08 -.03 -.26** -.19** -.01 -.06 .25** Victim blaming .06 -.22** -.00 -.07 -.03 -.02 -.05 -.03 .11* -.27** Credibility Counsel roles Punitive Corrupt auth Social distance Myth 1 b Myth 2c Myth 3 d Myth 4e 138 Victim blaming Credibility Counsel roles .05 Punitive -.07 -.41** Corrupt auth .05 -.36** .38** Social distance -.14* -.30* .27** .28** Myth 1 .02 .09 -.05 -.13* -.09 Myth 2 -.16** -.09 -.01 -.09 .10* .07 Myth 3 -.12* -.04 .06 -.09 .07 .10 .26** Myth 4 .02 -.04 -.06 -.05 .06 .18** -.01 .06 Victim blaming -.26** -.15** .24** .02 .13* .05 .32** .30** .01 a Attitudes toward homosexuality, b most raped men are very upset by the incident, c it is impossible for a man to rape a man, d most raped men do not need counseling after the incident, e even a big, strong man can be raped by another man,* p <.05, ** p <.01

PAGE 139

Table 5-3. Stepwise li near regression predicting victim blaming Steps (N = 304 a) b (S.E.) t Step 1 Sex -.46 (.12) -.22 -3.90** Step 2 Sex -.51 (.12) -.24 -4.52** Count .00 (.00) .13 2.22* Step 3 Sex -.15 (.13) -.07 -1.08 Count .00 (.00) .18 3.25** Homosexuality attitudes -.24 (.05) -.33 -5.15** Step 4 Sex -.12 (.13) -.06 -0.94 Count .00 (.00) .17 3.20** Homosexuality attitudes -.21 (.05) -.30 -4.61** Credibility -.26 (.07) -.19 -3.47** Step 5 Sex -.14 (.13) -.07 -1.10 Count .00 (.00) .17 3.12** Homosexuality attitudes -.19 (.05) -.26 -4.17** Credibility -.22 (.07) -.17 -3.13** Counseling role -.05 (.05) -.06 -1.00 Punitive orientation .18 (.05) .19 3.37** Social distance .07 (.06) .06 1.09 Step 6 Sex -.11 (.12) -.05 -0.90 Count .00 (.00) .17 3.41** Homosexuality attitudes -.16 (.04) -.23 -3.83** Credibility -.17 (.07) -.13 -2.56** Counseling role -.04 (.04) -.05 -0.54 Punitive orientation .17 (.05) .19 3.57** Social distance .03 (.06) .03 0.54 Rape myth 2 .21 (.04) .24 4.54** Rape myth 3 .16 (.04) .18 3.61** R2 change, step 1 .05** R2 change, step 2 .02* R2 change, step 3 .08** R2 change, step 4 .03** R2 change, step 5 .06** R2 change, step 6 .11** Model R2 .34 Model R2 adj .32 Model F 16.97** Model df 9, 294 a Incomplete cases were excluded through listwise deletion, p <.05, **p <.01. 139

PAGE 140

Table 5-4. Inmate cred ibility paired sample t -tests Mean a (SD) Lower Upper t-value b Muscular inmateHomosexual inmate 3.34 (1.03) 3.76 (0.88) -.52 -.33 -8.61** Muscular inmateYoung inmate 3.34 (1.03) 4.08 (0.65) -.83 -.65 -15.56** Muscular inmateGang member 3.34 (1.03) 3.34 (1.05) -.08 .07 -0.13 Muscular inmateInmate who owes money 3.34 (1.03) 3.77 (0.83) -.52 -.33 -9.04** Muscular inmateInmate who is drunk or high on drugs 3.34 (1.03) 3.28 (0.98) -.03 .15 1.31 Muscular inmateDelayed reporting the incident 3.34 (1.03) 3.52 (0.89) -.28 -.08 -3.65** Muscular inmatePreviously consented to sex with inmates 3.34 (1.03) 3.28 (1.00) -.04 .15 1.14 Muscular inmateHas mental health problems 3.34 (1.03) 3.48 (0.88) -.24 -.04 -2.70** Homosexual inmate Young inmate 3.76 (0.88) 4.08 (0.65) -.40 -.24 -7.58** Homosexual inmate Gang member 3.76 (0.88) 3.34 (1.05) .32 .51 8.45** Homosexual inmate Inmate who owes money 3.76 (0.88) 3.77 (0.83) -.09 .08 -0.11 Homosexual inmate Inmate who is drunk or high on drugs 3.76 (0.88) 3.28 (0.98) .39 .58 10.05** Homosexual inmate Delayed reporting the incident 3.76 (0.88) 3.52 (0.89) .15 .33 5.33** Homosexual inmate Previously consented to sex with inmates 3.76 (0.88) 3.28 (1.00) .38 .57 10.09** Homosexual inmate Has mental health problems 3.76 (0.88) 3.48 (0.88) .18 .38 5.52** Young inmate Gang member 4.08 (0.65) 3.34 (1.05) .64 .83 15.35** Young inmate Inmate who owes money 4.08 (0.65) 3.77 (0.83) .24 .38 8.91** Young inmate Inmate who is drunk or high on drugs 4.08 (0.65) 3.28 (0.98) .71 .89 17.22** Young inmate Delayed reporting the incident 4.08 (0.65) 3.52 (0.89) .47 .64 13.14** Young inmate Previously consented to sex with inmates 4.08 (0.65) 3.28 (1.00) .70 .89 16.12** 140

PAGE 141

141 Table 5-4. Continued. Mean a (SD) Lower Upper t-value b Young inmate Has mental health problems 4.08 (0.65) 3.48 (0.88) .52 .68 14.07** Gang memberInmate who owes money 3.34 (1.05) 3.77 (0.83) -.51 -.33 -9.15** Gang memberInmate who is drunk or high on drugs 3.34 (1.05) 3.28 (0.98) -.02 .15 1.49 Gang memberDelayed reporting the incident 3.34 (1.05) 3.52 (0.89) -.27 -.08 -3.66** Gang memberPreviously consented 3.34 (1.05) 3.28 (1.00) -.03 .15 1.29 Gang memberHas mental health problems 3.34 (1.05) 3.48 (0.88) -.24 -.03 -2.57* Inmate who owes money Inmate who is drunk or high on drugs 3.77 (0.83) 3.28 (0.98) .40 .58 10.98** Inmate who owes money Delayed reporting the incident 3.77 (0.83) 3.52 (0.89) .15 .34 5.31** Inmate who owes money Previously consented to sex with inmates 3.77 (0.83) 3.28 (1.00) .39 .58 9.89** Inmate who owes money Has mental health problems 3.77 (0.83) 3.48 (0.88) .20 .38 6.20** Inmate who is drunk or high on drugsDelayed reporting the incident 3.28 (0.98) 3.52 (0.89) -.33 -.16 -5.64** Inmate who is drunk or high on drugs Previously consented 3.28 (0.98) 3.28 (1.00) -.09 .08 -0.12 Inmate who is drunk or high on drugsHas mental health problems 3.28 (0.98) 3.48 (0.88) -.29 -.18 -4.69** Delayed reporting the incidentPreviously consented to sex with inmates 3.52 (0.89) 3.28 (1.00) .16 .31 6.20** Delayed reporting the incident Mental health problems 3.52 (0.89) 3.48 (0.88) -.04 .12 1.00 Previously consented to sex with inmatesHas mental health problems 3.28 (1.00) 3.48 (0.88) -.29 -.11 -4.22** a Reverse-coded item, 1 = always, 2 = generally, 3 = sometimes, 4 = rarely, 5 = never, b Mean difference tested through paired t -test, p <.05, ** p <.01

PAGE 142

Table 5-5. Inmate credibili ty zero-order correlations Education Sex Age Race Religiosity Seniority Satisfaction Stress Count ATH a Credibility Education Sex -.01 Age .16** -.10 Race .06 .11* -.04 Religiosity -.03 .06 .18** .25** Seniority .16** -.11* .52** .06 .15** Satisfaction -.04 -.01 .10* -.03 .19** .00 Stress .02 .07 .13* .00 .08 .15** -.15** Count .11* .16** .05 -.06 -.07 -.03 .05 -.13* ATH .04 .51** -.08 -.03 -.26** -.19** -.01 -.06 .25** Credibility -.02 .19** .03 -.04 .06 -.00 .10 .11* .00 .18** Victim b lamin g Counsel r oles Punitive Corrupt auth Social distance Myth 1 b Myth 2c Myth 3 d Myth 4eCredibility Victim b lam i ng 142 Counsel roles -.15** Punitive .24** -.39** Corrupt auth .02 -.36** .38** Social distance .13* -.29** .27** .28** Myth 1 .05 .09 .01 .09 -.10 Myth 2 .32** -.16** -.09 -.01 -.09 .10 Myth 3 .30** -.05 .05 -.09 .07 -.26** .26** Myth 4 .01 -.04 -.06 -.05 .06 .01 -.01 .06 Credibility -.26** .05 -.07 .05 -.14* .02 -.16** -.12* .02 a Attitudes toward homosexuality; b most raped men are very upset by the incident; c it is impossible for a man to rape a man; d most raped men do not need counseling after the incident; e even a big, strong man can be raped by another man;* p <.05, ** p <.01

PAGE 143

Table 5-6. Stepwise lin ear regression predicting inmate credibility Full sample (N = 375) Males (N = 260) Females (N = 115) Steps b (S.E.) t b (S.E.) t b (S.E.) t Step 1 Sex .28 (.08) .19 3.69** Step 2 Sex .27 (.08) .18 3.55* Stress .03 (.02) .11 2.06* .04 (.02) .12 2.01* .02 (.03) .06 0.68 Step 3 Sex .20 (.09) .13 2.28* Stress .04 (.02) .11 2.22* .04 (.02) .14 2.31* .02 (.03) .06 0.66 Homosexuality attitudes .06 (.03) .11 1.88 .11 (.04) .17 2.80** -.02 (.05) -.05 -0.49 Step 4 Sex .16 (.09) .10 1.86 Stress .03 (.02) .11 2.13* .04 (.02) .14 2.32* .00 (.03) .01 0.09 Homosexuality attitudes .03 (.03) .06 1.07 .09 (.04) .14 2.27* -.07 (.05) -.13 -1.46 Victim blaming -.16 (.04) -.21 -4.15** -.10 (.04) -.15 -2.50* -.39 (.09) -.41 -4.58** Step 5 Sex .17 (.09) .11 2.03* Stress .04 (.02) .10 2.20* .05 (.02) .14 2.37* 143 .00 (.03) .01 0.13 Homosexuality attitudes .03 (.03) .05 0.87 .08 (.04) .13 2.12* -.07 (.05) -.14 -1.59 Victim blaming -.15 (.04) -.20 -3.88** -.09 (.04) -.14 -2.24* -.39 (.09) -.40 -4.52** Social distance -.09 (.04) -.11 -2.20* -. 09 (.05) -.11 -1.76 -.10 (.07) -.12 -1.42

PAGE 144

144 Table 5-6. Continued Full sample (N = 375) Males (N = 260) Females (N = 115) b (S.E.) t b (S.E.) t b (S.E.) t Step 6 Sex .17 (.09) .11 1.99* Stress .04 (.02) .11 2.17* .04 (.02) .14 2.26* .00 (.03) .01 0.13 Homosexuality attitudes .03 (.03) .05 0.87 .09 (.04) .14 2.33* -.07 (.05) -.14 -1.66 Victim blaming -.13 (.04) -.18 -3.23** -.08 (.04) -.11 -1.74 -.33 (.09) -.35 -3.69** Social distance -.08 (.04) -.11 -2.11* -. 08 (.05) -.11 -1.73 -.11 (.07) -.15 -1.70 Rape myth 2 -.04 (.03) -.06 -1.20 -.09 (.04) -.15 -2.38* .06 (.05) .10 1.15 Rape myth 3 -.01 (.03) -.01 -0.16 .06 (.04) .10 1.61 -.16 (.06) -.24 -2.61* R2 change, step 1 .04** R2 change, step 2 .01* .02* .00 R2 change, step 3 .01 .03** .00 R2 change, step 4 .04** .02* .16** R2 change, step 5 .01* .01 .02 R2 change, step 6 .00 .03* .05* Model R2 .11 .10 .23 Model R2 adj .10 .08 .19 Model F 6.65** 4.85** 5.49** Model df 7, 368 6, 253 6, 108 p <.05, ** p <.01

PAGE 145

CHAPTER 6 DEFINITIONS AND WILLINGNESS TO RESPOND QUANTITATIVE RESULTS Introduction This chapter reviews the remaining two resear ch questions regarding correctional officer attitudes about sexual assault and sexual assault vi ctims in jail. These two questions, outlined in Chapters 1 and 3 examine how officers define ra pe in a jail atmosphere and whether certain contextual factors influence their definitions. The other research que stion explores officer willingness to respond to consensual sex and sexual assault between inmates. A paired sample t test will first examine whether significant differences exist across types of definitions and response. Second, I examine bivariate correl ations between the dependent variables and demographic, work, and independent variables to determine if there are significant associations. This chapter also discusses the results of for ced-entry stepwise ordi nary-least squares (OLS) regression models to determine which demogra phic and independent variables are important predictors of definitions of rape and willingne ss to respond. Finally, the analysis of willingness to respond is separated by educat ion level, where applicable, to determine whether the effects of the predictor variables vary across level of education. Rape Definitions The third research question of this study explores how officers define rape in a jail atmosphere. Specifically, it examines their attitudes regarding sexual consent after verbal and physical threats or in exchange for goods or prot ection. Similar to the an alysis of the previous two dependent variable s, a paired samples t -test is conducted first to determine if there are significant differences in how officers define rape. Second, the resu lts of zero-order correlations are presented to determine what variables are si gnificantly associated with rape definitions. Finally, variables that are significantly correlated with rape definitions are entered in a forced145

PAGE 146

entry stepwise regression model to determine their relative contribution to predicting definitions of rape. Definitions and Context Recall from Chapter 3 that there are five vi gnettes in the survey describing different scenarios between two inmates. Tw o scenarios depict situations in which an inmate has sex with another inmate because he is either physically overpowered or is physic ally threatened with death. A third situation describes an inmate havi ng sex with another inmate after the inmate is verbally threatened with being labeled an informant. A fourth scenario stat es that an inmate who is an informant has sex with another inmate in exchange for protection from other inmates. The last scenario states that an inmate has sex with another inmate to pay off a debt for goods. These last two scenarios describe situa tions that are typically referred to as quid pro quo situations, where inmates have sex in exchange for favors or protection (C honco, 1989; Fishman, 1934; Weiss & Friar, 1974). It is important to note, howe ver, that there is some coercion or threat in these last two scenarios as one inmate demands e ither sex in return for protection or threatens a physical assault in payback for borrowed money or goods. Respondents were asked to indicate their level of agreement regarding whether each s cenario depicts rape. The sample means range between 4.45 and 5.16 (where a resp onse of six indicates agreement that the vignette depicts rape and a response of one suggests it does not depict rape), suggesting that officers tend to view these scenarios as involving rape, consistent with prior research in a pris on setting (Table 6-1) (Eigenberg, 2000a). Paired sample t -tests are used to examine whether the mean differences between each scenario are significant. In ot her words, are officer levels of agreement about whether the scenario depicts rape signifi cantly different across scenarios? Although prior research has descriptively examined the tendency of officers to define these scenarios as rape, it has failed to 146

PAGE 147

determine whether mean differences in defining th ese types of scenarios as rape are significant (Eigenberg, 2000a). Paired sample t -tests are used becau se each respondent indicated a response for each scenario; therefore, responses for one scenario may be dependent on responses for other scenarios (Agresti & Finlay, 1997). The results of the paired sample t -tests indicate that there are significant mean differences regard ing several rape definition scen arios (Table 6-1). Recall that these items are coded so that higher values repr esent more agreement with defining the scenario as rape. A response of implies that the respo ndent strongly agrees that the scenario depicts rape, while a response of indicates that they strongly disagree that the scenario depicts rape. The sample mean for the scenario that depi cts an inmate physically overpowering another inmate is significantly higher than the sample me an for three other scenar ios (Table 6-1). There is no significant mean difference between the ph ysical overpowering scenario and the scenario where one inmate threatens to kill another unless they submit to sex. The scenario where an inmate threatens to kill another inmate unless th ey submit to sex is also significantly more likely to be defined as rape than situ ations involving threats to label an inmate as an informant or exchanges for protection or goods. These fi ndings are similar to and build upon findings by Eigenberg (2000a); specifically, o fficers are rather clear in defi ning situations as rape when inmates are either physically forced or threatened with violence. Officers in this sample are least likely to define quid pro quo situati ons as rape than any other situation, and they do not significantly diff erentiate between the quid pro quo situations. In other words, there is no significant difference be tween mean scores in the scenario involving provision of protection and the s cenario involving payment for money or goods in exchange for sex. The sample is less consistent in defining as rape the scenario depicting a verbal threat to 147

PAGE 148

label an inmate as an informant. When compar ed to physical overpowering or physical threats of murder, this item is significantly less likely to be defined as rape. Compared to quid pro quo situations, however, the verbal threat scenario is significantly more likely to be defined as rape. The results imply that officers are significantly mo re likely to see situat ions involving physical force or physical threats of murder as rape than situations involving threats to label, or quid pro quo situations (i.e., situ ations where an inmate provides protection or money or goods in exchange for sex). Overall, variation in coer cion and context appear to make a difference in applying the rape definition for respondents in this sample. This finding is somewhat consistent with the literature on rape victims in the community as police officers also use situational cues and context (like use of force or victim behavior) to define inci dents as rape or victims as credible (Amir, 1971; Campbell & Johnson, 1997; Feldman-Summers & Palmer, 1980; Weis & Borges, 1975). Bivariate Correlations An examination of prior studies revealed that although not much research has been conducted in this area, one study contributes to knowledge about what factors may be correlated with and predict how officers define rape. Eigenberg (2000a) found that those who avoid blaming victims are more likely to define assaultive situations as rape. She also examined how attitudes about professional orie ntation affected definitions of rape and found that preference for social distance and concern over co rruption of authority affected definitions of rape directly (neither counseling nor punitive orientations signifi cantly predicted officer definitions of rape in her study). Officers who preferre d less social distance and those more concerned with corruption of authority had less restrictive definitions about rape (Eigenbe rg, 2000a). Therefore, victim blaming and the four professional orientation items are entered into a correlation analysis with definitions of rape. 148

PAGE 149

Other variables entered in the bivariate analys is include sex, age, race, attitudes toward homosexuality, perceptions of inma te credibility, and male rape myths. Disbelief in rape myths was important in predicting inclusive definitions of rape in other a prior study (Jenkins & Dambrot, 1987). Seniority, job satisfaction, stre ss, and facility size (inmate count) are also included in the correlation analys is. Many of these demographic and work variables are included as they are significant predictors of officer attitudes in other research (Arthur, 1994; Burt, 1980; Farkas, 1999; Jimenez & Abreu, 2003; Jurik, 1985; Lee & Cheung, 1991; Mori et al., 1995; Nagel et al., 2005; Poole & Regoli, 1980a; Sh eldon & Parent, 2002; Teske & Williamson, 1979; Xenos & Smith, 2001). Examination of the zero-order correlations s hows that none of the following variables is significantly correlated with the rape definitions index: sex, at titudes toward homosexuality, victim blaming, counseling roles, punitive orientation, corruption of authority, social distance, male rape myths, seniority, jobs satisfaction, stress or facility size (Table 6-2). These results are somewhat surprising, as a prior study of a prison sample found that vic tim blaming, corruption of authority, and social distance were significant predictors of whethe r officers defined a situation as rape (Eigenberg, 2000a). Because there ar e no significant differences between males and females or officers with or without a college de gree for any of the individual rape definition items or the index, it was not su rprising that these two variables were not bivariately correlated with definitions of rape (Table 4-3). Tw o other demographic variables, race (r = -.11, p <.05) and age (r = .13, p<.05), are significantly associated with th e rape definition index, however (Table 6-2). Defining scenarios as rape is significantly associated with being White and being older. In addition, the perception of inmate re ports of rape as credible is also significantly associated (r = .26, p<.001) with defining a situation as rape. 149

PAGE 150

Multivariate Analysis Because none of the variables found important in previous research are significantly correlated with definitions of rape in this study, the order of vari ables entered into the stepwise models is neither theoretical no r based on prior research. Instea d, the variables are entered in a fashion similar to previous analyses, and demogr aphic variables were entered in the first steps1. Age is entered as the first variable in a stepwise regression model, followed by race and perceptions of inmate credibility (Table 6-3). Age adds little explained variance in the dependent variable in this model (1%). Age is significant, however; older respondents are more willing to define these incidents as depicting rape In the next step, race also adds only 1% of explained variance to the model and is insignificant. Age remains significant in this step. Lastly, perception of inmate credibility is entered to examine its unique effect after controlling for demographic factors. Perception of inmate credib ility is significant and explains an additional 6% of the variance in definitions of rape. Respondents who perceive inmates who report rape as credible are significantly more lik ely to define these incidents as rape. Age remains significant in this step, and race remains insignificant as a pred ictor of definitions of rape across all steps of the model. Overall, the model explains a relatively sma ll portion (8%) of variance in definitions of rape among this sample, indicating that other factor s not included in this study are more important in predicting this dependent variable. Examination of the standardized coefficients and R2 change across steps indicates that perceptions of inmate credibility is the strongest predictor of whether officers defi ne these incidents as rape. Following this, age is the second 1 As noted in chapter 3, the regression model for definitions of rape is estimated using robust standard errors to correct for the problem of heteroscedasticity. 150

PAGE 151

most powerful predictor in this model2. The effects of age are not unexpected, given that prior research on officer attitudes indicates that ol der officers have more positive attitudes about rehabilitation, human service ro les, and inmates in general th an younger officers (Cullen et al., 1989; Farkas, 1999; Griffin, 2002; Jurik, 1985; Kifer et al., 2003; Klofas, 1986; Maahs and Pratt, 2001; Toch & Klofas, 1982; Teske & W illiamson, 1979; Whitehead et al., 1987). Summary of Rape Definition Analysis The results of this section indicate that offi cers do distinguish cont extual differences in defining certain sexual incidents between inmates as rape. They are significantly more likely to define instances involving physical force or thre ats of violence as rape than incidents involving simple verbal threats or quid pro quo situa tions. Consistent with the literature on nonincarcerated female victims of rape and police o fficer perceptions, correctional officers also use contextual cues when defining instances as rape In the multivariate analysis, older respondents and those who perceive inmates as credible are much more likely to define the situations in the scenarios as rape. Similar to the victim blaming results, perceptions of inmate credibility is an important factor in predicting wh ether officers define situations as rape. The next section will discuss willingness to respond to acts of consen sual sex and sexual assault among this sample. Willingness to Respond The fourth research question in this study examines how officers perceive their own willingness to respond to acts of consensual sex a nd sexual assault in jail. Recall there are three indices that measure officer willingness to respo nd to these incidents in jail (Table 3-5). For each of the items in these indices, I report the results of paired sample t -tests to examine 2 Results were estimated using both listwise deletion an d expectation-maximization (EM) imputation methods to address missing values in incomplete cases. The results we re not substantively different using either method. Since the true values of the missing data are unknown, the results presented in Table 6-3 use listwise deletion to address missing values. 151

PAGE 152

significant differences in how officers are willing to respond to incidents of consensual sex and sexual assault. Following this, I report bivariate correlations to indicate which variables are significantly associated with each willingness to respond index. The variables that are significantly associated with will ingness to respond are then entere d into separate theoretically driven forced-entry stepwise regression models to determine relative significance of each of the variables in predicting each willingn ess to respond dependent variable. Differences in Response Type The first step of this analysis compares the sample means of each of the eight individual willingness to respond items to determine if officers are significantly more willing to respond in certain ways to consensual sex and sexual assa ult between inmates. Although the willingness to respond dependent variables were indexed accordi ng to the results of the factor analysis discussed in Chapter 3, the t -tests comparing differences betw een individual items are intended to offer additional insight about which responses officers are most comf ortable with. It is apparent from this analysis that although offi cers do conceptually view responding by talking to inmates and using proactive measures as separate concepts, they also be lieve that they should respond differently to consensual sex versus sexual assault. Recall that response values for these items range from one through six, and each item was reverse coded so higher values reflect greater agreement with the statement about wi llingness to respond. For example, a response value of reflects the answer option strongly disagree while a response of indicates strongly agree that officers should respond in this way (Tables 4-3). Responding to consensual sexual acts First, I will discuss the results of the paired sample t -tests for willingness to respond regarding consensual sexual acts between inma tes. I will then discuss results for officer 152

PAGE 153

willingness to respond regarding instances of sexual assault. Following this, I will compare the means for the similarly worded response items for both consensual se x and sexual assault. The mean for encouraging inmates to report c onsensual sexual acts is significantly lower than the mean encouraging officers to do everything they can to prevent consensual sex between inmates. In other words, officers have significantly lower levels of agreement for encouraging inmates to report consensual se xual acts than for believing officers should do everything to prevent consensual sex. Officers had a signifi cantly higher mean, however, for encouraging inmate reporting of consensual se x acts than for talking to inma tes about these acts. In other words, they are significantly more willing to encourage inmates to report consensual sex than they are to talk to inmates about consensual sex. Comparison of the mean sample responses for the item stating officers should do everything they can to prevent consensual sexual acts indicates that officers have higher mean levels of agreement with this response than w ith responses regarding talking to inmates about consensual sex. This result may indicate that officers are collectively less likely to support specific response items (like talking) than they are a vague and general description of officer response (like doing everything they can). Ov erall, however, it appe ars that regarding consensual sexual acts, officers are willing to encourage inmates to re port and are willing to do everything they can to prevent them. Officers ha ve significantly lower levels of willingness to talk to inmates to discourage consensual sexual ac ts. This indicates that officers are much less comfortable adopting this response when compared to the other responses. Responding to sexual assault The following section will discuss mean sample responses to the items about willingness to respond to instances of sexual assault. Officers have higher levels of agreement regarding doing everything they can to prevent sexual assault over encouraging inmate reporting of sexual 153

PAGE 154

assault. Though this suggests that officers may acknowledge their responsi bility over that of inmates to address these serious situations, the means for using proactive measures such as cell assignments or protective custody are significantly lower than the means for both of these items. In other words, they have greater willingne ss to respond to sexual a ssault by doing everything they can to prevent it and by advising inmates to report it than they do for using proactive measures to prevent it. Again, collective agreem ent to general statements about the nature of their response concerning sexual assault may be more agreeable for this sample than detailed statements about response type. Officers appear very willing to indicate that they should do everything they can to prevent sexual assaults in jail. The sample has significantly higher levels of agreement for this statement when compared to all other statements about willingness to respond. The sample mean is much higher for most other items in comparison with talking to inmates about the risk of sexual assault, however. Overall, this suggests officers are hesitant to discuss sexual assault risks with inmates, though they are even more hesitant to talk to inmates a bout consensual sex. They also have significantly higher levels of agreement for using cell assignments than for talking to inmates about the risk of sexual as sault or using protective custody to safeguard inmates. It appears then that they are willing to use the proa ctive measure of cell assignments than other methods to combat these instances in jail. Regarding t ype of proactive response, cell assignments appear preferable to protective custody, suggesting that officers may perceive that protective custody is useful for functions other than addressing sexual assault among inmates. The sample indicates a significantly lower mean and less agreement with using cell assignments than with encouraging inmates to report sexual as sault, and doing everythi ng they can to prevent sexual assaults. The sample mean was signifi cantly lower for usi ng protective custody to 154

PAGE 155

safeguard inmates than for doing everything to pr event sexual assault, again suggesting that officers are more supportive of general responses than specific measures. They have higher levels of agreement, however, for using protec tive custody than for talk ing to inmates about consensual sex or the ri sks of sexual assault. Comparison of responses for consensual sex and sexual assault Officers have significantly higher means and le vels of agreement for encouraging inmates to report sexual assault over encouraging them to report consensual sex. Officers also indicate significantly higher means for doing everything they can to prevent sexual assault over doing everything they can to prevent consensual sex. Officers also report sign ificantly higher means and more agreement for being willing to talk to inmates about the risks of sexual assault compared to talking to them about consensual se x. These results suggest that officers perceive the relative seriousness of sexual assault compared to consensual sex, are more willing to address sexual assault, and recognize their responsibility a nd the utility of proactive measures to combat the issue. Overall, officers are willing to respond to c onsensual sexual acts by encouraging inmates to report them and doing everything they can to prev ent them, however they not as willing to talk to inmates about these incidents. Regarding sexu al assault, officers appear more willing to use specific proactive measures like cell assignments and protective custody than in just talking to inmates about the risk of sexual assault. They are more willing to use proactive measure of cell assignments than many other methods to combat these instances in ja il. Though they have significantly higher levels of agreement for usi ng protective custody than for talking to inmates about the risks of sexual assault, they appear to prefer cell as signments to proactively combat these situations. Similar to consensual sexual acts, the sample is more willing to respond to sexual assault by doing everything they can to prevent it than by advising inmates to report it. 155

PAGE 156

However, they are more willing to advise inmates to report sexual assault than they are to use proactive measures to prevent it. Though they wa nt to do everything they can to prevent sexual assault, they are also hesitant to talk to inmates about the risks of sexual assault. Overall, summarizing the t-test results for the indices and the individual items, officers are most willing to do everything to prevent either cons ensual sexual acts or sexual assault compared to encouraging inmates to report these situations (Table 6-4). They appear least comfortable with talking to inmates about th ese instances. These results could be caused by the nature of their job, as officers perceive it as their responsibility to ensure th at the rules of the institution are implemented and followed. They may be implying their willingness to do their job when they respond that officers should be willing to respond to rule violations (inc luding sexual assault and consensual sex). They may also, however, be collectively less willing to support specific response items (i.e., talking to in mates) than they are a vague and general description of officer response (i.e. doing everything they can). In addition, if officers adhere to the belief that their job is not to counsel inmates, bu t to enforce rules and manage in mate behavior, they may believe it is outside the real m of their job respons ibilities to discuss these sens itive matters with inmates. They also consistently indicate that they pe rception of sexual assault as more serious by indicating significantly higher willingness to res pond to these incidents than consensual sex incidents. Bivariate Correlations Prior research suggests that in predicting willingness to respond to prison rape, the religious, officers with counseli ng orientations, officers who pref er less social distance from inmates, and officers who condemn homosexuali ty are more willing to respond (Eigenberg, 1994). Therefore, religiosity, attitudes toward homosexuality, and the professional orientation items (counseling roles, punitive orientations, corr uption of authority, and social distance) are 156

PAGE 157

incorporated into a zero-order correlation analysis (Table 6-5). Education level is entered in the correlation analysis as the results discussed in Chap ter 4 indicate that at least for one index, there may be a significant effect of education on willingness to respond. In addition, age, race, sex, seniority, job satisfaction, and stress are included in the correlation analysis as other research indicate s their importance in predicting officer attitudes (Arthur, 1994; Burt, 1980; Farkas, 1999; Ji menez & Abreu, 2003; Jurik, 1985; Lee & Cheung, 1991; Mori et al., 1995; Nagel et al., 2005; Poole & Regoli, 1980a; Sheldon & Parent, 2002; Teske & Williamson, 1979; Xenos & Smith, 2001). Facility size (inmate count) is also included in the correlation analysis. Perceptions of credibi lity is included in the analysis to determine if whether officers believe inmate ra pe reports affects their willingness to respond. Male rape myth items are entered to examine whether these vari ables are associated with willingness to respond to sexual incidents between inmates. Only variables significantly associated with at least one willingness to respond index is presented in the bivariate correlation table (Table 6-5). Education, relig iosity, age, attitudes toward homosexuality, perceptions of credibility corruption of authority, and two rape myth items are significantly associated with willi ngness to respond by encouraging inmate reporting and officer prevention of consensual sex and se xual assault (Table 6-5). Willingness to respond in this way is significantly associ ated with no college degree (r = -.11, p <.05), religiousness (r = .15, p<.01) and older respondents (r = .11, p<.05). This dependent variab le is also significantly correlated with disapproval of homosexuality (r = -.16, p<.01), belief in credibility of inmates who report rape (r = .21, p<.01), concern with corruption of authority (r = .19, p<.01), agreeing that other men cannot rape men (r = -.14, p<.01), and agreeing that men do not need counseling after rape incidents (r = -.15, p<.01). 157

PAGE 158

The index measuring officer willingness to re spond by talking to inmates about consensual sex and about the risk of sexual as sault is significantly associated with religiosity, age, seniority, perceptions of credibility, couns eling roles, and social distance. Specifically, willingness to respond in this way is correlate d with being re ligious (r = .11, p<.05), being older (r = .17, p<.05), and having more experience in corrections (r = .18, p <.01). This dependent variable is also significantly correlated with believ ing inmate reports of rape (r = .14, p<.05), accepting counseling roles (r = .19, p<.01), and preferring less social distance from inmates (r = .19, p<.01). Lastly, the index that indicates officer willingness to respond to sexual assault by using proactive measures such as cell assignments a nd protective custody is si gnificantly associated with education, age, job satisfac tion, perceptions of cred ibility, social distance and two male rape myth items. Willingness to respond using proactiv e action is significantly related to having a college degree (r = .14, p <.01), being older (r = .21, p<.01), being satisfied with the job (r = .11, p<.05), believing inmates who report rape (r = .22, p<.01), and preferring less social distance from inmates (r = .26, p <.01). This dependent variable is also significantly correlated with disagreement with the myths that men cannot be raped (r = .22, p<.01) and do not need counseling after a rape incident (r = .17, p<.01). Multivariate Analysis This section reports the findings of theoretica lly driven forced-entry stepwise regressions for each willingness to respond dependent variable Only variables that are significantly associated with the willingness to respond indices are incorporated in the multivariate analysis. The variables are ordered in a manner similar to previous analyses; demographic and work variables are incorporated first, followed by attitudes toward homose xuality, perception of inmate credibility, professional orientation items, and male rape myth items. The use of this order allows an examination of the specific contri bution of variables entered later such as male 158

PAGE 159

rape myth items after controlling for variables th at are significant predictors of willingness to respond in prior research. W illingness to respond by encouraging inmate reporting and officer prevention practices is analyzed separately by level of education as the results presented in Chapter 4 suggest that response type may vary significantly by whether officers have a college degree or not3. Encouraging inmate reporting and prevention-full sample The variables that are significantly correlated with this wi llingness to respond index are entered in a theoretically driven stepwise regr ession model with seven equations. Demographic variables and those that have si gnificantly predicted willingness to respond in prior research are entered in the initial steps in order to determin e the unique contribution of variables entered in the final step. The first equation examines th e effect of education on willingness to respond by encouraging inmate reporting and officer preven tion. Though this variable adds only about 1% of explained variance to the model, it is sign ificant and suggests that those without a college degree are more likely to indicate that they w ould respond to sexual assa ult and consensual sex between inmates in this manner. The second step examines the impact of relig iosity on willingness to respond. Those who report higher levels of religios ity are significantly more likely to report a willingness to respond to consensual sex and sexual assault between inmates by enc ouraging inmate reporting and officer prevention practices (Table 6-6). Religiosity contributes about 2% of explained variance to the dependent variable. In step three, age adds only about 1% of explained variance to the dependent variable and is insignificant. Educat ion and religiosity maintain significance in this 3 Although chapter 4 results also indicate significant diff erences in willingness to respond using proactive measures for officers with and without a college degree, separation of the sample by level of education did not reveal any substantive differences in parameter estimates. In addition, coefficient comparison tests ( z-tests) revealed no significant differences in the effects of the independent variables across level of education. 159

PAGE 160

equation. In step four, attitudes toward homosexua lity adds another 2% of explained variance to willingness to respond in this way. This variable is also significant, and those who report less tolerance of homosexual lifesty les are more likely to be wi lling to respond by encouraging inmate reporting and officer prevention practices, similar to prior research (Eigenberg, 1994). Both religiosity and age are insignificant in this step; however, educati on remains significant. Step five incorporates perceptions of inmate credibility, which adds 5% of explained variance to the model. This vari able is also significant; officers who perceive inmates reports of rape as credible are also si gnificantly more likely to respond by encouraging inmate reporting and officer prevention practices. Education and attitudes toward homosexuality remain significant in this step and religiosity and age re main insignificant. Corruption of authority is added to the model in step six and explains an additional 4% of variance. Those who are concerned with corruption of aut hority are significantly more likel y to report being willing to respond to consensual sex and sexua l assault in this way. This is an interesting finding as prior research suggests that other professional orie ntation items are more important in examining willingness to respond (Eigenberg, 1994). However, for this sample, officers who perceive that it is possible for inmates to manipulate and corrup t their authority are more willing to encourage inmates to report consensual sex acts and sexual assault and to believe that officers should do everything they can to prevent these incidents. Though age and religiosity remain insignificant in this step, education, attitude s toward homosexuality, and perceptions of inmate credibility remain significant. Step seven incorporates the two male rape myth s that are bivariately associated with this willingness to respond index. Together these variables add 3% of explained variance to the dependent variable; however, only one rape myth is significant. Those w ho adhere to the male 160

PAGE 161

rape myth that raped men do not need counseling are significantly less likely to report that they are willing to respond to sex and sexual assa ult among inmates in this manner. Education, attitudes toward homosexuality, perceptions of inmate credibili ty, and corruption of authority maintain significance in this equation, while religiosity and age remain insignificant. About 18% of the variance in willingness to respond by encouraging inmate reporting and officer prevention is explained by this model. Examination of R2 changes suggests that perceptions of inmate credibility adds the most explained variance to this model, followed by corruption of authority and male rape myths. The standardized coefficients across steps also suggest that corruption of aut hority and perceptions of inmate credibility are the strongest predictors in the model, and th at attitudes toward homosexuality is also a relatively strong predictor4. Officers in this sample who are willing to respond by encouraging inmate reporting and officer prevention practices are those who are concerned about th e corruption of their authority, perceive inmate reports of rape as cr edible, do not adhere to the myth that raped men do not need counseling, and who condemn homosexuality. In addition, officers who do not have a college degree are also more w illing to reply in this manner. The models were analyzed separately by e ducation because the results from Chapter 4 indicate that those w ith and without a college degree are significantly different in their willingness to respond to one of the items in th is index (Table 4-4). Specifically, responses differ significantly for the item stating that offi cers should do everything they can to prevent consensual sexual acts in jail; indicating that officers with no degree have higher levels of 4 Results were estimated using both listwise deletion an d expectation-maximization (EM) imputation methods to address missing values in incomplete cases. Although th e results did not change dramatically, the analysis with imputed data indicates age as a significant predictor of willingness to respond in this way for the full sample. Because the true values of the missing data are unknown, howeve r, the results presented in Table 6-6 use listwise deletion to address missing values. 161

PAGE 162

support for this statement. In addition, the results for the full sample suggest significant differences in willingness to respond between these two groups. No college degree sample. The next part of the analysis separated the samples by level of education, and therefore the stepwise models for both officers with and without a college degree incorporate only five steps (step one, previous ly incorporating the variable education, is eliminated). The first step for the sample of officers without a college degree incorporates religiosity and contributes only 1% of explained variance (Table 66). In this step and later steps, religiosity is insignificant. The next st ep (step three for the full sample) adds age to the model. Similar to religiosity, ag e adds very little explained variance to the model (less than 1%) and is insignificant in both this and later equations. The addition of attitudes toward homosexuality (step four for the full sample) also adds little explained variance to the dependent variable (2%) and is in significant this step. The next step (step five for the full sample) incorporates perceptions of inmate credibility; this variable adds 3% of explai ned variance to the model. Offi cers without a college degree who perceive inmate reports of rape as credible are more willing to respond in this manner. Attitudes toward homosexuality becomes significant in th is step, indicating that officers who condemn homosexuality are more likely to respond in this way. The next step (step six for the full sample) includes corruption of authority and adds 7% of explained variance in willingness to respond. This variable is also significant, and indicates that officers concerned with the corruption of their authority are more willing to respond by encour aging inmate reporting and officer prevention practices. Perceptions of inmate credibility main tains significance in this step, however attitudes toward homosexuality once again becomes insignifi cant. Officer concerns over the corruption of their authority may be a more important factor than their attitudes about homosexuality in 162

PAGE 163

explaining this dependent variable The final step of the model (step seven for the full sample) adds 5% of explained variance in the dependent variable and indicates th at officers who adhere to the myth that raped men do not need counselin g after the incident are significantly less likely to be willing to respond by encouraging inmate reporting and officer prevention practices. Attitudes toward homosexuality, perceptions of in mate credibility, and corruption of authority are also significant in this final equation. The total variance in willi ngness to respond by encouraging inmate reporting and officer prevention explained by this model is about 18 percent. Examination of R2 change and standardized coefficients suggests that for officers without a college degree, corruption of authority is the most powerful predictor of willingness to respond by encouraging inmate reporting and officer prevention practices. Attitu des about male rape myth and perceptions of inmate credibility are also relatively important predictors5. Officers without a college degree who are concerned with the corruption of their au thority, who believe inmate reports of rape, and who dismiss the myth that raped men do not need counseling are more likely to agree that officers should both encourage inmate reporting a nd should do everything they can to prevent consensual sex and sexual assault among inmates. College degree sample. The first step for the sample of officers with a college degree (step two for the full sample) indicates that religi osity is not a significant predictor of willingness to respond in this manner for these officers. This variable also adds relatively little explained variance to the model (2%). Incl uding age in the model (step three for the full sample) explains an additional 5% of variance in the dependent va riable. Age is significant; older officers are 5 Results were estimated using both listwise deletion an d expectation-maximization (EM) imputation methods to address missing values in incomplete cases. The results we re no differences in results using either method. Since the true values of the missing data are unknown, the results presented in Table 6-6 use listwise deletion to address missing values. 163

PAGE 164

more likely to indicate willingness to res pond by encouraging inmate reporting and officer prevention practices. Age is also significant in al l remaining steps of this model. The next step (step four for the full sample) indicates that th e addition of attitudes toward homosexuality does not explain much more variance in the model (1%). This variable is also not a significant predictor of willingness to respond. The next step (step five for the full sample) incorporates perceptions of inmate credibility and suggests that officers with a college degree who are concerned about the corruption of their authority are more willing to respond by encour aging inmate reporting and officer prevention practices. This variable adds 7% of explained variance to the mode l. The next step (step six for the full sample) adds corruption of authority to the model but does not increase explained variance by much (2%). This variable is also insignificant here and in the fi nal step. In the final step (step seven for the full sample), the addi tion of the two rape myth items add 2% of explained variance to the model, though neither is a significant predictor of willingness to respond in this way. Similar to the previous two models, about 18% of the total varian ce in willingness to respond by encouraging inmate reporting and o fficer prevention is explained by this Examination of R2 change and standardized coefficients in the model for officers with a college degree indicates that perceptions of credibility is the most important variable across all steps6. Officers with a degree who believe inmate repor ts of rape are significantly more likely to indicate willingness to respond by en couraging inmate reporting and officer prevention practices. 6 Results were estimated using both listwise deletion an d expectation-maximization (EM) imputation methods to address missing values in incomplete cases. The results we re not substantively different using either method. Since the true values of the missing data are unknown, the results presented in Table 6-6 use listwise deletion to address missing values 164

PAGE 165

Age is also a powerful predictor across steps and implies that older officers are also quite willing to respond by encouraging inmate repor ting and officer pr evention practices. The full sample results imply that for the entire sample, perceptions of inmate credibility and concern with corruption of authority is a powerful predictor of being willing to respond. Officers who tend to dismiss the myth that raped men do not need counseling are also significantly more likely to respond by encouraging inmate reporting and officer prevention practices. The results for the full sample also suggest that officers without a college degree may also be more willing to respond in this way. In analyzing those with and without a college degree separately, it appe ars that the results for officers without a college degree are somewhat similar to the full sample, which could result from the number of officers withou t a degree (n = 219) compared to the number of officers with a degree (n = 156). Attitudinal fact ors matter more for the full sample and for officers without a college degree. Officers without a degree who ar e concerned with corrup tion of authority, who perceive inmates as credible, and who dismiss the myth that raped men do not need counseling are more likely to indicate wi llingness to respond to consen sual sex and sexual assault by encouraging inmate reporting a nd officer prevention practices. Although other variables are relevant in the fu ll sample, perceptions of inmate credibility and concern over corruption of auth ority are two of the strongest predictors in both samples. Though perceptions of inmate credibility is also the most important predictor for officers with a college degree, the demographic variable of age is also a powerful predictor of this dependent variable. For officers with a de gree, believing inmate accounts of rape and being older increases the likelihood of being willi ng to respond by encouraging in mate reporting and officer prevention practices. Overall, perceptions th at inmates who report rape are credible is 165

PAGE 166

consistently important in predicting officer willingness to respond by encouraging inmate reporting and officer prevention Ot her variables that are important factors in predicting whether officers agree that they should respond to se x among inmates in this way are concern over corruption of authority, belief in male rape myth, and age. Talking to inmates The following variables are significantly asso ciated with the willingness to respond by talking to inmate index: religiosity, age, seniority, perceptions of inmate credibility, counseling roles, corruption of authority, and social di stance. These variables are entered into a theoretically driven forced-entry stepwise regr ession model to determine the relative significance of each variable and to examine coefficient a nd explained variance change. The first step includes religiosity and only expl ains 1% of the total variance in willingness to respond by talking to inmates (Table 6-7). This variable is not a significant predictor of willingness to respond by talking to inmates. The second step in corporates seniority and its inclusion explains an additional 3% of the variance in the dependent vari able. Officers with more experience are significantly more likely to report th at they would be willing to ta lk to inmates about consensual sexual acts and the risks of sexual assault. Step three incorporates age but adds little explained variance to the model (1%). In addition, age is not a significant predicto r of willingness to talk to inmates about these issues. In step four, th e addition of perceptions of inmate credibility explains an additional 2% of va riance in the dependent variable Those who perceive inmate reports of rape as credible are also significan tly more willing to respond by talking to inmates about these incidents. Incorporating counseling roles into the fifth equation adds an additional 3% of explained variance to the model. Officers who embrace couns eling roles are also significantly more likely to indicate that they would respond in this manner. This finding is similar to what Eigenberg 166

PAGE 167

(1994) found in her study of officers in a prison sample. Seniority and perceptions of inmate credibility are also still significant in this equation while religiosity is still insignificant. In the final step, the variable social dist ance is also significant, but only adds 2% of explained variance to the model. Officers who prefer social dist ance in their relationships with inmates are significantly less likely to indicate that they would be willing to talk to inmates about consensual sex among inmates and about the risk of sexual a ssault; again, a finding similar to previous research (Eigenberg, 1994). The model explains about 12% of the total variance in willingness to respond by talking to inmates. Examination of R2 change and standardized coefficien ts suggests that seniority is the most powerful predictor of wh ether officers are willing to respond to sexual assault and consensual sex between in mates by talking to them7. Officers who have more experience are willing to respond in this way. Perceptions of inmate credibility, counseling roles and social distance also matter; those who believe inmate reports of rape, embrace counseling as part of their job, and prefer less social distance in thei r relationships with inmates are more willing to respond to these incidents by talking to inmates. This finding is consistent with Eigenbergs (1994) findings about the importance of counseling roles and social distance. Proactive measures The next part of the analysis tests the re lative importance of independent variables on officer willingness to respond by using proactiv e measure such as cell assignments and protective custody to safeguard inmates form sexual assault. Only variable s that are significantly associated with willingness to respond using proa ctive measures to safegu ard inmates are entered 7 Results were estimated using both listwise deletion an d expectation-maximization (EM) imputation methods to address missing values in incomplete cases. The results we re no differences in results using either method. Since the true values of the missing data are unknown, the results presented in Table 6-7 use listwise deletion to address missing values. 167

PAGE 168

into the stepwise model8. The model consists of six steps a nd incorporates education in the first step. Age is added in step two, followed by job satisfaction, perceptions of inmate credibility, social distance, and male rape myths. Educational level is a significant predicto r of willingness to respond by using proactive measures in step one of the stepwise model (T able 6-8). Officers with a college degree are significantly more likely to indicate willingness to respond by using proactive measures to safeguard inmates form sexual assa ult. Education explains 2% of the variance in willingness to respond using proactive measures. In step two, ag e is added and explains an additional 4% of variance in the dependent variable. Age is al so significant; older officers are more willing to respond to sexual assault by using proactive measures to safeguard inmates. Incorporating job satisfaction in step three adds only 1% of explained variance to the model; this variable is also not significant in this or later steps. Step four includes perceptions of inmate credibility and adds 4% of explained va riance to the dependent variable. Officers who perceive inmates as credible are also significantly more willing to use proactive measures to protect inmates from sexual assault. Social dist ance explains an additional 3% of variance in the dependent variable when it is adde d to step five. Officers who pref er less social distance in their relationships with inmates are significantly more likely to indicate that they would respond in this manner. This finding is consistent with results from the analysis of willingness to respond by talking to inmates and implies officers who pref er social distance in their relationships with inmates are less likely to be willing to take pr oactive measures like talking with inmates and acting to safeguard them from sexual assault. 8 As noted in chapter 3, one of the regression models in this section (willingness to respond using proactive measures ) is estimated using robust standard erro rs to correct for the probl em of heteroscedasticity. 168

PAGE 169

Male rape myths are added to the model in step six and incorporate an additional 3% of explained variance in the depende nt variable. Education become s insignificant in this equation, however, age, perception of credibility, and social distance rema in significance in this step. Neither male rape myth is significant, suggest ing that beliefs about male rape do not act as important predictors of whether officers are willing to use proactive measures to respond. The model explains about 17% of the total variance in willingness to respond by using proactive measures. Examination of R2 change and standardized coe fficients suggests that social distance, perceptions of inmate credibility, and age are the most powerful predictors in this model9. Officers who prefer less soci al distance in their relationshi ps with inmates, who believe that inmate reports of rape are credible, a nd who are older are significantly more likely to indicate that they would respond in this manner, similar to prev ious research (Eigenberg, 1994). In other words, officers who agree that they sh ould have compassion and act as advocates for inmates are more willing to discuss the risks of sexual assault with them and to use jail administrative procedures to safeguard them. Summary of Willingness to Respond Analysis According to the paired sample t -tests, officers indicate they are most willing to do everything to prevent either consensual sexual ac ts or sexual assault compared to encouraging inmates to report these acts. They appear far less comfortable talk ing to inmates about consensual sex and the risk of sexual assault. They are more willing to advise inmates to report sexual assault than they are to use proactive me asures to prevent it. Though they are willing to 9 Results were estimated using both listwise deletion an d expectation-maximization (EM) imputation methods to address missing values in incomplete cases. Although th e results did not change dramatically, the analysis with imputed data indicates male rape myths are significant predictors of willingness to respond proactively. Because the true values of the missing data are unknown, however, the results presented in Table 5-3 use listwise deletion to address missing values 169

PAGE 170

do everything they can and specifically use proactiv e measures to prevent sexual assault, they are hesitant to talk to inmates about the risks of se xual assault. In predicting willingness to respond to these incidents, demographic and work variables, attitudes about homosexuality, per ceptions of inmate credibility, professional orientation items, and male rape myth items appear important; but they vary in their e ffect on these dependent variables. For encouraging inmates to report and officers to prevent these incidents, results for the full sample and for officers without a coll ege degree are very similar and suggest that perceptions of inmate credibi lity, concern over corruption of au thority, and beliefs about male rape myths are important. For officers with a co llege degree, perceptions of credibility also matters, but age is also an important factor. Ov erall, officers who perceive inmate reports as credible, regardless of educational background, are willing to respond to consensual sex and sexual assault by encouraging inmate reporting and officer prevention. In contrast, seniority and profe ssional orientation are factors that are more important when it comes to talking to inmates about these incide nts. Officers with more experience, those who accept counseling roles as part of their job function, and those who prefer less social distance in relationships with inmates are significantly more willing to talk to inmates about consensual sex among inmates and about the risk of sexual assault. This finding is similar to results of prior research (Eigenberg, 1994) and suggests that of ficers with experience working with inmates, and who embrace the human service function of the j ob are more open to discussing these sensitive topics with inmates. There is a relatively st rong correlation in this an alysis between age and seniority (Table 6-5); given this, the results of seniority are not su rprising. Prior research about officer attitudes suggest that ol der officers are typically more pos itive in their attitudes toward inmates, rehabilitation, and huma n service roles of their job (C ullen et al., 1989; Farkas, 1999; 170

PAGE 171

Griffin, 2002; Jurik, 1985; Kifer et al., 2003; Klofas, 1986; Maahs & Pratt, 2001; Toch & Klofas, 1982; Teske & Williamson, 1979; Whitehead et al., 1987). Preference for less social distance in relationships with inmates also significantly predicts a willingness to respond to these incidents proa ctively by safeguarding inmates with cell assignments and protective custody. The results for the social distance variable indicates that officers concerned with compassion and advocacy fo r inmates are more willing to take proactive measures like talking with and safeguarding th em (compared to simply encouraging inmate reporting and generally believing th at officers should do what they can to prevent these acts). Older officers and those who perceive inmate report s of rape as credible are also more likely to respond in this manner. Across the multivariate models predicting the willingness to respond indices, perceptions of inmate credibility is cons istently important in predicti ng officer willingness to respond to consensual sex and sexual assault. Officers who perceive inmate re ports of rape as credible are much more willing to respond to these incident s by encouraging inmate reporting and officer prevention, by talking to inmates, and by using pr oactive measures to protect inmates. The effects of education across all wi llingness to respond indices are not as consistent, however. While results for the full sample suggest that o fficers without a college de gree are significantly more likely to respond to these incidents by enco uraging inmate reporting and officer prevention practices, officers with a college degree are more likely to use proactive measures to combat sexual assault in jail. Regarding professional orientation items, they appear relevant, yet how they affect each dependent variable differs across models. Officer s concerned with corrup tion of their authority are more likely to respond to these incident s by encouraging inmate reporting and officer 171

PAGE 172

prevention practices. In contra st, acceptance of counseling role s and less preference for social distance in relationships with in mates predicts willingness to talk to inmates about consensual sex and the risk of sexual assault. Social di stance is also import ant regarding proactive measures; officers who prefer less social distance are also more willing to respond proactively to sexual assault in jail by using cell assignments and protective custody. Conclusion In defining rape among inmates in jail, officer s in this sample are most likely to view situations involving physica l overpowering or physical threats as rape. Though they also tend to perceive instances involving verb al threats as rape, they have higher levels of agreement for those instances including physical force or threats. They are l east likely to define as rape situations that involved coerci on or quid pro quo, where inmates submitted to sex in exchange for goods or protection. These findings are suppo rtive of those in st udies examining police officer perception of rape among non-incarcerated females that indicate that context of the incident matters in defining it as rape (Amir, 1971; Campbell & Johnson, 1997; FeldmanSummers & Palmer, 1980; Weis & Borges, 1975). Multivariate analysis of the rape definitions index suggest that older officers and those who perceive inmate re ports of rape as credible are more likely to define situations as rape. In addressing sexual assault and consensual sex among inmates, officers in this sample have high levels of agreement that they should do everything they can to prevent such instances. They are also willing to suggest that officers s hould encourage inmates to report these instances. They have lower levels of agreement for using proactive measures to combat sexual assault and safeguard inmates, however. Finally, they are leas t willing to talk to inmates about consensual sex and the risk of se xual assault in jail. 172

PAGE 173

The multivariate analysis of the willingness to respond indices reveal several important findings. Perceptions of inmate credibility, prof essional orientation items, and male rape myth items appear important; but how they affect each willingness to respond index varies. Officers who perceive inmate reports of rape as credible are consistently more willing to respond to acts of consensual sex and sexual assau lt. Overall, there is variation in which professional orientation items significantly influence which willingness to respond index. Corruption of authority is an important factor in predicti ng whether officers indicate will ingness to respond by encouraging inmate reporting and prevention practices. C oncern over corruption of authority predicts willingness to respond in this way for the full model and for officers without a college degree. For willingness to talk to inmates, however acceptance of counseling roles and social distance are the professional orie ntation items that have significant influence. Officers who accept counseling roles as part of their job function and who prefer less social distance in their relationships with inmates are more likely to indi cate willingness to respond in this way. This finding is similar to results of prior research (Eigenberg, 1994) and suggests that officers who may have spent more time around inmates or simply have more life experience, as well as those who embrace the human service job function are mo re willing to discuss these sensitive topics with inmates. Social distance is also impor tant in predicting willingness to respond by using proactive measures as well. Preference for less social dist ance in relationships with inmates significantly predicts willingness to respond by using cell assi gnments and protective custody to safeguard inmates from sexual assault. The results for the social distance variable indicate that officers concerned with compassion and advocacy for inmate s are more willing to take specific actions such as talking with and safe guarding them (compared to enc ouraging reporting and believing 173

PAGE 174

that officers should do what they can to prevent these situations). Regarding the influence of male rape myth acceptance, officers in the full model and those without a college degree who disregard the myth that raped men do not need counseling are also mo re likely to respond by encouraging inmate reporti ng and prevention practices. 174

PAGE 175

175 Table 6-1. Rape defini tions paired sample t -tests Mean a (SD) Lower Upper t-value b Physically overpowersThreatens label informant 5.16 (0.97) 4.73 (1.32) .30 .56 6.60** Physically overpowersThreatens kill 5.16 (0.97) 4.73 (1.32) -.03 .20 1.41 Physically overpowersProvides protection 5.16 (0.97) 4.47 (1.39) .54 .83 9.33** Physically overpowersLoans goods 5.16 (0.97) 4.45 (1.39) .44 .74 7.79** Threatens label informant Threatens kill 4.73 (1.32) 5.07 (1.08) -.44 -.25 -7.28** Threatens label informant Provides protection 4.73 (1.32) 4.47 (1.39) .15 .36 4.65** Threatens label informant Loans goods 4.73 (1.32) 4.45 (1.39) .05 .28 2.81** Threatens kill Provides protection 5.07 (1.08) 4.47 (1.39) .49 .72 10.52** Threatens kill Loans goods 5.07 (1.08) 4.45 (1.39) .40 .62 9.21** Provides protection Loans goods 4.73 (1.32) 4.45 (1.39) -.19 .00 -1.89 a Reverse-coded item, 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongly disagree, b Mean difference tested through paired t -test, p <.05, **p <.01

PAGE 176

Table 6-2. Definitions of rape zero-order correlations Education Sex Age Race Seniority Satisfaction Stress Count ATH a V ictim b laming Definitions Education Sex -.01 Age .16** -.10 Race .06 .11* -.04 Seniority .16** -.11* .52** .06 Satisfaction -.04 -.01 .10* -.03 .00 Stress .02 .07 .13* .00 .15** -.15** Count .11* .16** .05 -.06 -.03 .05 -.13* ATH .04 .51** -.08 -.03 -.19** -.01 -.06 .25 Victim blaming .06 -.22** -.00 -.07 -.02 -.05 -.03 .11* -.30** Definitions .03 .06 .13* -.11* .08 .06 .07 .04 .04 -.08 176 Credibility Counsel r oles Punitive Corrupt auth Social distance Myth 1 b Myth 2c Myth 3 d Myth 4eDefinitions Credibility Counsel roles .05 Punitive .07 -.40** Corrupt auth .05 -.36** .38** Social distance -.14* -.29** .27** .28** Myth 1 .02 .09 -.05 -.13* -.09 Myth 2 -.16** -.09 -.01 -.09 .10 .07 Myth 3 -.12** -.05 .05 -.09 .07 .10 .26** Myth 4 .02 -.04 -.06 -.05 .06 .18** -.01 .06 Definitions .26** -.02 .04 .08 -.04 -.05 .01 -.07 -.06 a Attitudes toward homosexuality, b most raped men are very upset by the incident, c it is impossible for a man to rape a man, d most raped men do not need counseling after the incident, e even a big, strong man can be raped by another man;* p <.05, ** p <.01 176

PAGE 177

Table 6-3. Stepwise linear regression predicting definitions of rape Steps (N = 335a) b (S.E.) t Step 1 Age .07 (.02) .10 2.67** Step 2 Age .06 (.02) .10 2.37* Race -.29 (.16) -.12 -1.78 Step 3 Age .05 (.02) .10 2.01* Race -.27 (.16) -.11 -1.66 Credibility .34 (.08) .24 4.26** R2 change, step 1 .01* R2 change, step 2 .01** R2 change, step 3 .06** Model R2 .08 Model R2 adj .07 Model F 9.58** Model df 3, 331 a Incomplete cases were excluded through listwise deletion, p <.05, ** p <.01 177

PAGE 178

Table 6-4. Willingness to respond paired sample t -tests Mean a (SD) Lower Upper t-value b Indices Respond prevent report indexRespond talk index 5.22 (0.04) 4.29 (0.06)0.80 1.06 14.12** Respond prevent report indexRespond proactive index 5.22 (0.04) 4.82 (0.97)0.31 0.50 8.20** Respond talk indexRespond proactive index 4.29 (0.06) 4.82 (0.97)-0.66 -0.39 -7.44** Individual Items Encourage inmate report consensualEncourage inmate report sexual assault 4.60 (1.39) 5.48 (0.71) -1.02 -0.75 -12.73** Encourage inmate report consensualEverything prevent consensual 4.60 (1.39) 5.22 (1.00) -0.75 -0.49 -9.48** Encourage inmate report consensualEverything prevent sexual assault 4.60 (1.39) 5.60 (0.62) -1.15 -0.86 -13.81** Encourage inmate report consensualTalk to inmates consensual 4.60 (1.39) 4.18 (1.49) 0.24 0.59 4.64** Encourage inmate report consensualTalk to inmates sexual assault 4.60 (1.39) 4.41 (1.41) -0.01 0.37 1.90 Encourage inmate report consensualUse cell assignments 4.60 (1.39) 4.92 (1.03) -0.48 -0.16 -3.88** Encourage inmate report consensualUse protective custody 4.60 (1.39) 4.73 (1.19) -0.31 0.03 -1.62 Encourage inmate report sexual assaultEverything prevent consensual 5.48 (0.71) 5.22 (1.00) 0.16 0.36 5.08** Encourage inmate report sexual assaultEverything prevent sexual assault 5.48 (0.71) 5.60 (0.62) -0.17 -0.07 -4.51** Encourage inmate report sexual assaultTalk to inmates consensual 5.48 (0.71) 4.18 (1.49) 1.13 1.45 16.12** Encourage inmate report sexual assaultTalk to inmates sexual assault 5.48 (0.71) 4.41 (1.41) 0.92 1.21 13.93** Encourage inmate report sexual assaultUse cell assignments 5.48 (0.71) 4.92 (1.03) 0.45 0.66 10.24** Encourage inmate report sexual assaultUse protective custody 5.48 (0.71) 4.73 (1.19) 0.62 0.87 12.06** Everything prevent consensualEverything prevent sexual assault 5.22 (1.00) 5.60 (0.62) -0.46 -0.29 -8.45** Everything prevent consensualTalk to inmates consensual 5.22 (1.00) 4.18 (1.49) 0.88 1.19 13.17** Everything prevent consensualTalk to inmates sexual assault 5.22 (1.00) 4.41 (1.41) 0.65 0.97 9.86** 178

PAGE 179

179 Table 6-4. Continued Mean a (SD) Lower Upper t-value b Everything prevent consensualUse cell assignments 5.22 (1.00) 4.92 (1.03) 0.16 0.43 4.29** Everything prevent consensualUse protective custody 5.22 (1.00) 4.73 (1.19) 0.35 0.62 7.02** Everything prevent sexual assaultTalk to inmates consensual 5.60 (0.62) 4.18 (1.49) 1.25 1.56 18.13** Everything prevent sexual assaultTalk to inmates sexual assault 5.60 (0.62) 4.41 (1.41) 1.04 1.33 15.93** Everything prevent sexual assaultUse cell assignments 5.60 (0.62) 4.92 (1.03) 0.57 0.78 12.82** Everything prevent sexual assaultUse protective custody 5.60 (0.62) 4.73 (1.19) 0.75 0.99 14.15** Talk to inmates consensual Talk to inmates sexual assault 4.18 (1.49) 4.41 (1.41) -0.35 -0.11 -3.79** Talk to inmates consensual Use cell assignments 4.18 (1.49) 4.92 (1.03) -0.89 -0.56 -8.67** Talk to inmates consensual Use protective custody 4.18 (1.49) 4.73 (1.19) -0.72 -0.37 -6.12** Talk to inmates sexual assault Use cell assignments 4.41 (1.41) 4.92 (1.03) -0.65 -0.34 -6.37** Talk to inmates sexual assault Use protective custody 4.41 (1.41) 4.73 (1.19) -0.47 -0.14 -3.70** Use cell assignments Use protective custody 4.92 (1.03) 4.73 (1.19) 0.07 0.30 3.20** a Reverse-coded item, 1 = strongly agree, 2 = agree, 3 = not sure, probably agree, 4 = not sure, probably disagree, 5 = disagree, 6 = strongly disagree, b Mean difference tested through paired t -test, p <.05, **p <.01

PAGE 180

Table 6-5. Willingness to respond zero-order correlations Education Sex Age Race Religiosity Seniority Satisfaction Stress A TH a Respond 1 b Respond 2cRespond 3 d Education Sex -.01 Age .16** -.10 Race .06 .11* -.04 Religiosity -.03 .06 .18** .25** Seniority .16** -.11* .52** .06 .15** Satisfaction -.04 -.01 .10 -.03 .19** .00 Stress .02 .07 .13* .00 .08 .15** -.15** ATH .04 .51** -.08 -.03 -.26** -.02 -.01 -.06 Respond 1 -.11* .02 .11* -.05 .15** .05 .08 .09 -.16** Respond 2 .06 .01 .17* .07 .11* .18** .05 .05 .01 .33** Respond 3 .14** .03 .21** -.04 -.00 .09 .11* .07 .05 .37** .31** Credibility C ounsel r oles Punitive Corrupt auth Social distance Myth 1eMyth 2f M yth 3g M yth 4hRespond 1 Respond 2 Respond 3 Credibility 180 Counsel roles .05 Punitive -.07 -.39** Corrupt auth .05 -.36** .38** Social -.14* -.29** .27** .28** Myth 1 .16** .09 .01 .09 -.10 Myth 2 -.16** -.09 -.09 -.09 .10 .10 Myth 3 -.12* -.05 .05 -.09 .07 -.26** .26** Myth 4 .02 -.04 -.06 -.05 .06 .01 -.01 .06 Respond 1 .21** .09 .04 .19** -.08 -.09 -.14** -.15** .05 Respond 2 .14* .19** -.02 -.09 -.19** .04 .02 -.02 .04 .33** Respond 3 .22** .04 .00 .03 .26** -.01 -.22** -.17** .08 .37** .31** a Attitudes toward homosexuality, b willing to respond by encourage in mate reporting/ officer prevention, c willing to talk to inmates, d willing to use proactive measures, e most raped men are very upset by the incident, f it is impossible for a man to rape a man, g most raped men do not need counseling after the incident, h even a big, strong man can be raped by another man, p <.05, ** p <.01

PAGE 181

Table 6-6. Stepwise linear regression predicting willingness to respond by encouraging inmate reporting & officer prevention Full sample (N = 313a) No college degree (N = 181) College degree (N = 132) Steps b (S.E.) t b (S.E.) t b (S.E.) t Step 1 Education -.17 (.08) -.12 -2.04* Step 2 Education -.16 (.08) -.11 -1.95 Religiosity .04 (.02) .13 2.39* .03 (.02) .12 1.62 .05 (.03) .17 1.73 Step 3 Education -.18 (.08) -.13 -2.22* Religiosity .04 (.02) .12 2.04* .03 (.02) .12 1.60 .03 (.03) .10 1.13 Age .04 (.02) .11 1.85 .00 (.03) .00 .03 .09 (.03) .22 2.54* Step 4 Education -.17 (.08) -.12 -2.09* Religiosity .03 (.02) .08 1.44 .02 (.02) .08 1.08 .02 (.03) .08 .85 Age .04 (.02) .10 1.78 .00 (.03) .01 .09 .08 (.03) .21 2.39* Homosexuality attitudes -.07 (.03) -.13 -2.18* -.07 (.04) -.14 -1.82 -.05 (.05) -.09 -1.05 Step 5 Education -.17 (.08) -.12 -2.09* Religiosity .02 (.02) .06 1.00 .02 (.02) .05 .69 .02 (.03) .06 .73 Age .03 (.02) .09 1.66 .00 (.03) .01 .09 .07 (.03) .18 2.15* Homosexuality attitudes -.09 (.03) -.17 -2.99** -.09 (.04) -.17 -2.25* -.08 (.05) -.16 -1.76 Credibility .22 (.06) .22 3.95** .17 (. 07) .17 2.32* 181 .27 (.09) .26 3.10**

PAGE 182

Table 6-6. Continued Full sample (N = 313a) No college degree (N = 181) College degree (N = 132) Steps b (S.E.) t b (S.E.) t b (S.E.) t Step 6 Education -.17 (.08) -.12 -2.22* Religiosity .02 (.02) .08 1.45 .02 (.02) .08 1.02 .03 (.03) .08 .96 Age .04 (.02) .09 1.74 .00 (.03) .01 .16 .07 (.03) .19 2.18* Homosexuality attitudes -.07 (.03) -.14 -2.44* -.07 (.04) -.13 -1.78 -.07 (.05) -.13 -1.47 Credibility .22 (.06) .20 3.74** .17 (. 07) .17 2.34* .25 (.09) .24 2.86** Corruption of authority .18 (.05) .21 3.83** .23 (.06) .26 3.38** .13 (.08) .15 1.74 Step 7 Education -.20 (.08) -.14 -2.60* Religiosity .02 (.02) .08 1.47 .03 (.02) .09 1.23 .02 (.03) .07 .85 Age .04 (.02) .10 1.76 .01 (.03) .02 .31 .07 (.03) .18 2.10* Homosexuality attitudes -.09 (.03) -.17 -3.02** -.08 (.04) -.17 -2.29* -.08 (.05) -.14 -1.59 Credibility .18 (.05) .18 3.38** .15 (.07) .15 2.08* .22 (.09) .22 2.55* Corruption of authority .16 (.05) .18 3.40** .21 (.06) .24 3.40** .12 (.08) .13 1.53 Male rape myth 2 -.06 (.03) -.10 -1.77 -.04 (.04) -.08 -1.12 -.10 (.07) -.13 -1.50 Male rape myth 3 -.08 (.04) -.14 -2.42* -.11 (.04) -.20 -2.77** -.02 (.07) -.03 -.30 R2 change, step 1 .01* R2 change, step 2 .02* .01 .02 R2 change, step 3 .01 .00 .05* R2 change, step 4 .02* .02 .01 R2 change, step 5 .05** .03* .07** R2 change, step 6 .04** .07** .02 R2 change, step 7 .03** .05** .02 Model R2 .18 .18 .18 Model R2 adj .15 .15 .14 Model F 8.10** 5.44** 3.93 Model df 8, 304 7, 173 7, 124 a Incomplete cases were excluded through listwise deletion p <.05, **p <.01 182

PAGE 183

Table 6-7. Stepwise linear regression predicting willingness to respond by talkin g to inmates Ste p s ( N = 332a ) b ( S.E. ) t Ste p 1 Religiosity .05 (.03) .09 1.72 Step 2 Religiosity .04 (.03) .07 1.33 Seniority .04 (.01) .19 3.40** Step 3 Religiosity .04 (.03) .06 1.15 Seniority .03 (.01) .13 2.11* Age .07 (.04) .11 1.67 Step 4 Religiosity .03 (.03) .05 0.95 Seniority .03 (.01) .14 2.22* Age .07 (.04) .10 1.55 Credibility .27 (.10) .14 2.66** Step 5 Religiosity .02 (.03) .03 .52 Seniority .03 (.01) .13 2.18* Age .07 (.04) .10 1.61 Credibility .25 (.10) .13 2.52* Counseling roles .18 (.06) .17 3.18** Step 6 Religiosity .01 (.03) .02 .37 Seniority .03 (.01) .15 2.38* Age .06 (.04) .08 1.34 Credibility .22 (.10) .12 2.23* Counseling roles .14 (.06) .13 2.40* Social distance -.22 (.09) -.14 -2.52* R2 change, step 1 .01 R2 change, step 2 .03** R2 change, step 3 .01 R2 change, step 4 .02** R2 change, step 5 .03** R2 change, step 6 .02* Model R2 .12 Model R2 adj .10 Model F 7.10** Model df 6, 325 a Incomplete cases were excluded through listwise deletion p <.05, ** p <.01 183

PAGE 184

184 Table 6-8. Stepwise linear regres sion predicting willingness to respond proactively Steps (N = 337a) b (S.E.) t Step 1 Education .27 (.10) .15 2.72** Step 2 Education .20 (.10) .12 2.06* Age .10 (.02) .18 3.97** Step 3 Education .21 (.10) .12 2.16* Age .09 (.02) .17 3.82** Satisfaction .04 (.02) .08 1.80 Step 4 Education .19 (.10) .12 1.91 Age .09 (.02) .17 3.69** Satisfaction .03 (.02) .06 1.26 Credibility .29 (.07) .21 4.39** Step 5 Education .21 (.10) .12 2.15* Age .08 (.02) .15 3.26** Satisfaction .03 (.02) .06 1.12 Credibility .25 (.07) .18 3.69** Social distance -.21 (.06) -.19 -3.59** Step 6 Education .19 (.10) .10 1.97 Age .08 (.02) .15 3.21** Satisfaction .03 (.02) .06 1.09 Credibility .21 (.07) .15 2.94** Social distance -.20 (.06) -.18 -3.46** Male rape myth 2 -.11 (.06) -.14 -1.91 Male rape myth 3 -.08 (.05) -.09 -1.60 R2 change, step 1 .02** R2 change, step 2 .04** R2 change, step 3 .01 R2 change, step 4 .04** R2 change, step 5 .03** R2 change, step 6 .03** Model R2 .17 Model R2 adj .15 Model F 9.18** Model df 7, 329 a Incomplete cases were excluded through listwise deletion p <.05, **p <.01

PAGE 185

CHAPTER 7 PERCEPTIONS OF PROFESSIONAL ORIENTATION AND INMATE ATTITUDES QUATITATIVE RESULTS Introduction A second purpose of the study is to examine co rrectional officer professional orientations and their perceptions about the orientations of their fellow officers. Are correctional officers in jails in Florida accurate in their perceptions of how their fellow officers regard inmates and their profession? The research will examine how precis ely correctional officers identify their fellow officers beliefs regarding counseling roles, pun itive orientation, corrupt ion of authority, and social distance with inmates. A final purpose of this research is to explore how correctional officers perceive inmate attitudes about sexual assault in jails. Specifically, do correctional officers in jails perceive that inmates blam e victims and adhere to male rape myths? Perceptions of Fellow Officers Professional Orientation One finding in prior studies of correctional officers implies that officers believe their colleagues to be more pro-custody and opposed to non-custody work (such as counseling roles) than they actually are, a term re ferred to as pluralistic ignoran ce (Cullen et al ., 1989; Klofas & Toch, 1982). Pluralistic ignorance exists when members of a group have misconceptions about the beliefs of their fellow me mbers (Kauffman, 1981). In the first study to examine this phenomenon among correctional officers, Kauffman (1981) found that offi cers view their peers to be less favorable to treatment and more aversive to inmates than they actually are. Officers did not perceive their fellow offi cers as willing to act as advocates for inmates, however, officers in the sample were relatively willing to do so. Klofas and Toch (1982) found that officers in their sample generally underestimated the levels of professionalism amo ng their fellow officers and overestimated their preference for a custody or ientation and social distance from inmates. They also found that officers underestimated the nu mber of officers who feel that compassion for 185

PAGE 186

inmates is important. The following sections present findings about officer attitudes and their perceptions of those of their fellow officers toward professi onal orientation item such as counseling roles, punitive orientation, corr uption of authority, a nd social distance. Counseling Roles A visual inspection of percentages of the c ounseling role items indicates that 55% of officers in the sample agree that counseling roles should be left to mental health professionals (and that officers do not accept counseling roles as pa rt of their job) (Table 7-1). Most of the sample (73%) believes that their colleague s agree with this statement and do not accept counseling roles. The results are similar for the item stating, counseling is a job for counselors, not correctional officers. While 54% of the sample agrees with this item, 73% of officers believe their fellow officers would agree that counse ling is not a job for corre ctional officers. In addition, while 66% of officers think their fellow o fficers would agree that an officer wanting to do counseling should change jobs only 48% of officers actually agree with this item. These results imply that officers percei ve their peers to be more aversi ve to counseling roles that they are. The paired sample t -tests reveal significant differences between the attitudes that officers express and those they perceive their collea gues to have. The means for actual officer expressions of counseling role ite ms are significantly higher ( t -values of 5.94, p<.01, 7.59, p<.01, and 6.54, p<.01, respectively) than the means for thei r perceptions of the attitudes of their fellow officers. Recall that highe r response values for the counse ling roles items indicate greater acceptance of counseling roles. Thus, higher means for expressed attitudes indicate that they are significantly more accepting of counseling roles than they perceive themselves and their fellow officers to be. These finding suggest that officers in this sample are plural istically ignorant about whether counseling roles should be part of thei r job functions. They perceive that their 186

PAGE 187

colleagues are much less accepting of counseling ro les as part of the job function than they actually are. Though neither of the studies discussed above specifically explored acceptance of counseling roles, Kauffman (1981) found that o fficers underestimated support for fellow officers to apply for a treatment position. Officers in that sample, therefore also overestimated aversion to treatment and counseling functions of the co rrectional environment. This study uniquely contributes to research about correctional officers by examining their attitudes and their perceptions of those of their fellow officers regarding the role of counseling in their profession. Punitive Orientation A visual inspection of percentages of the punitive orientation item that states rehabilitation programs are a wast e of time and money indicates that only 44% of officers in the sample tend to have a punitive orientation and ag ree with this statement (Table 7-1). Most of the sample (68%), however believes that thei r colleagues agree with this statement and are punitive. While 38% of officers agree with th e statement improving jails for inmates makes them worse for officers, about 51% of officers perceive that their fellow officers agreed with this statement. The results for the item stati ng, a military regime is the best way of running a jail are similar. About 70% of the sample agrees with this item while 77% of officers indicated that most or almost all of their colleagues would agree. These percentages imply that officers assume their peers have more punitive correctional orientations th an they actually do. According to the paired sample t -tests, there are also sign ificant differences between officer attitudes and their percep tions of those of their fellow officers. The means for actual officer expressions of all three item are significantly lower ( t -values of -7.94, p<.01, -5.31, p<.01, and -3.04, p<.01, respectively) than the means for thei r perceptions of their fellow officer attitudes. Recall that higher means for these re sponses indicate more punitiveness regarding this item. In other words, officers are actually le ss punitive than they believe their group to be 187

PAGE 188

regarding the punitive orientati on items of rehabilitation programs, improvement of jails, and operations of jails. The findings in this sec tion indicate that, like prior research, officers misperceive their fellow officers to be pro-cust ody when they are not (Cullen et al., 1989; Kauffman, 1981; Klofas & Toch, 1982). The results imply that offi cers in this sample perceive their fellow officers to be more punitive than they actually are. Corruption of Authority About 88% of the officers in the sample agree with the statement you cant ever completely trust an inmate suggesting concern over an inmates ability to corrupt their authority (Table 7-1). When officers were asked to de termine about how many of their fellow officers agree with this statement, 95% of the officers in the sample indicated th at their colleagues would agree. The paired samples t -test also indicates that there are significant differences between officer attitudes and their perceptions of these attitudes regarding this item. The t -value (-4.07, p<.01) indicates that the means fo r perceptions of their fellow offi cer attitudes are higher than the means for their actual attitudes. Recall that th ese items are reverse coded so that higher response values indicate more concern over corruption of authority. Therefore, officers tend to believe that their colleagues are more concerned with trusting inmates than they actually are. There are also similarities be tween percentages of officers who agree with the statement you must keep conversations with inmates shor t and businesslike and officer perceptions of whether their fellow officers agree with this stat ement. About 83% of of ficers in this sample agree with this statement and 88% indicated th at their fellow officers would agree with this statement. There are significant differences be tween the means of these statements, however. The t -value (-2.51, p<.05) indicates that actual officer att itude means are significantly lower than their perceptions of their fellow o fficers attitudes. Because higher response values for this item imply more concern over corrupti on of authority, this finding indi cates that officers perceive 188

PAGE 189

their fellow officers to be more concerned with this statement (and corrup tion of their authority) than their actual attitudes suggest. Officers in the sample also have relatively high levels of agreement for the item a good principle is not to get close to inmates. A bout 93% of the sample agrees with this item, indicating concern over corruption of authority. Likewise, 95% of officers guessed that their colleagues would agree with this item. About 94% of officers ag ree with the statement that a personal relationship with an inmate invites co rruption, and a similar percentage of officers (95%) indicated that their fello w officers would agree with this statement. There are no significant differences between the means for th ese items, indicating that officers have an accurate conception of their fellow officers att itudes regarding getting close to inmates, and whether personal relationshi ps invite corruption. About 89% of officers agree with the final corr uption of authority statement: if an officer is lenient with inmates they will take advantage of him or her. Similarly, about 92% of officers indicated that their fellow officers would agree with this statement. There are no significant differences between officer actual attitudes and their pe rceptions of their fellow officer attitudes regarding this statement. This suggests that, si milar to the items about getting close to inmates and having a personal relationship with inmates, officers are accurate about the perceptions of their fellow officer attitudes regard ing being lenient with inmates. Overall, the findings indicate that for two of the five corrupt ion of authority items, officers perceive each other to be more concerned with the corruption of authority than they actually are. The findings of the paired sample t -tests for the other three items indicate no significant differences between actual attitudes and perceptions of attitudes, although the di rection of the relationship is consistent with the other items. Regarding the ability of inmate s to corrupt officer authority then, officers have 189

PAGE 190

only moderate levels of plura listic ignorance. This study offe rs a unique contribution to the study of officer attitudes regard ing corruption of their authority and their perceptions of the concern of their fellow officers over this issue. Social Distance About 57% of the sample agrees with the statement a correctional officer should work hard to earn trust from inmates (Table 7-1). Only about 34% of the samp le perceives that their fellow officers would agree with this statement, however. While 69% of officers in the sample agree with the statement that it is important fo r a correctional officer to have compassion, only 43% of officers perceive that th eir fellow officers would agree to this item. Only about 34% of officers in the sample agree with the statemen t that stated, you get to like the inmates you supervise. Again, fewer officers actually percei ve that their fellow officers would respond this way. Only 24% of the officers in this sample indicated that their fe llow officers would agree with this item. While 38% of officers in the sample indicated agreement with the statement sometimes a correctional officer should be an advocate for an inmate, only 21% of officers believed their fellow officers responded in this wa y. The results for the statement the way to get respect from inmates is to take an interest in them are similar to those discussed above. While 29% of officers indicated that they agr ee with this statement, only about 23% of the sample indicated they believed their fellow officers would agree with it. All the means comparisons suggest that officers perceive that their fellow o fficers have more preference for social distance than they actually express. Paired sample t -tests reveal significant differences between the attitudes that officers express and those they perceive their fellow officers to have. All sample means for officer perceptions of their fellow officer attitudes regarding social di stance are significantly higher (tvalues of -7.77, p<.01, -4.84, p<.01, -3.67, p<.01, -5.99, p<.01, and -2.21, p<.05, respectively) 190

PAGE 191

than actual officer attitudes. Recall that highe r means for these items indicate more preference for social distance in relationships with inmates. The findings indicate that officers in the sample perceive each other to prefer more social distance in relationships with inmates than they actually prefer. Prior research also finds that officers tend to overestimate their preference for social distance and to underestimate their willingness to advocate and act compassionately for inmates (Kauffman, 1981; Klofas & Toch, 1982). Overall, officers in this sample are pluralisti cally ignorant with regard to their perceptions of their fellow officer attitudes. Officers in the sample are more accepting of counseling roles as part of their job function, are less punitive, and prefer less social distance in their relationships with inmates than they perceived themselves to be. They are also moderately less concerned over the ability of inmates to corrupt their aut hority than they perceived. These findings are consistent with prior studies that find officer s to be more accepting of rehabilitation, less custody oriented, and less aversive to inmates than they think they are (Cullen et al., 1989; Kauffman, 1981; Klofas & Toch, 1982). Perceptions of Inmate Attitudes This study offers a unique contribution to unde rstanding correctional officer perceptions of inmates that no other study has examined. Officer s were asked to indicat e their perceptions of inmate attitudes regarding victim blaming and male rape myth items. Specifically, officers were asked to estimate how many inmates in their fac ility would blame inmate victims of rape and would accept male rape myths. These results ar e then compared to th eir own attitudes about victim blaming and male rape myths. Victim Blaming About 18% of officers replied th at they agree that inmates w ho previously consent to sex with other inmates in jail get what they deserv e if they are subsequently raped (Table 7-2). 191

PAGE 192

Many more officers (64%) perceive that inmates would agree with this item, however. This finding is consistent among the remaining victim blame items. Most officers perceive that inmates in their facility will ag ree with blaming these victims, while the percentage of officers who agree with blaming these victims is comp aratively lower. Rega rding blaming inmates because of the way they act, only 10% of offi cers agree that these inmates deserved blame. About 61% of officers perceive th at inmates in their facility w ould agree with this item. Only 16% of the sample blames inmates who take mone y or cigarettes in exchange for sex, while 63% of the officers felt inmates in thei r facility would blame in this s ituation. Few officers (16% and 11%, respectively) agree with blaming inmates who take money or cigarettes for sex and inmates who act feminine; a greater percen tage (67% and 70%, respectively) of officers perceive that the inmates in their facility woul d blame these inmates, however. Similarly, only 10% and 11% of officers agree with blaming men who are not more careful or who do not fight back (respectively), while 57% of officers estimate that inmates in their facility would agree with these items. The paired sample t -tests reveal that for all blaming items, there are significant differences between officer beliefs and officer perceptions of inmate beliefs. The t -values are all negative, which indicates that the means for officer percep tions of inmate attitudes are higher than the means for officer attitudes. R ecall that higher response values correspond with blaming inmate victims for rape. Therefore, comparing the m eans for officers who blame victims to the means for officers who believe inmates will blame vict ims, officers perceive the inmate tendency to blame as significantly greater than their own tend ency to blame. Percentages of officers who blame victims are relatively low; the results suggest that they expect most of the inmates in their facilities to blame inmate victims of rape. 192

PAGE 193

Male Rape Myths Examination of the percentages indicates that 94% of officers agree with the male rape myth that men who are raped by other men are ups et by the incident. A smaller percentage of officers (87%) believe that most or all inmates in their facility would agree with this item. Similarly, 88% of officers believe that even a big, strong man can be raped by another man. Only 71% of the officers in this sample believe that the majority of inmates in their facility would agree. These results imply that officers are not prone to accept myths surrounding male rape, and they do not perceive that inmates in thei r facility are likely to accept these myths either. The analysis of the two remaining male rape myth items also suggests that officers generally do not accept and do not expect inmates in their facility to accept the myths surrounding male rape. Only 7% of officers agreed that it is impossible to rape a man, and only 8% agreed that men who are raped do not need counseling after the in cident. Although the sample was more willing to suggest that inmates ag ree with these rape myths, they did not assign majority percentages to their expectations of inmate agreem ent (21% and 34% respectively). There are significant differences between officer beliefs and officer perceptions of inmate beliefs regarding male rape myth acceptance. The negative t -values indicate that the means for officer perceptions of inmate attitudes are higher than the means for their own attitudes. Recall that higher response values indicat e acceptance of male rape myths. In comparing the mean of officers who accept male rape myths to the mean believing inmates accept them, officers perceive inmates to have greater acceptance of male rape myths over their own acceptance of these myths. The percentages of officers who ac cept rape myths are relatively low, implying that they expect most of the inmates in th eir facilities to accept male rape myths. 193

PAGE 194

194 Conclusion Officers in this sample are pl uralistically ignorant about thei r perceptions of their fellow officer attitudes. They are much more willi ng to accept counseling roles as part of their job duties and prefer less social dist ance from inmates than they pe rceive of themselves and fellow officers. They are also less punitive than they expect. They are modera tely less concerned over the ability of inmates to corrupt their authority than they percei ve. These results support prior research that implies that officers accept rehabi litation, are not custody oriented, and are not as anti-inmates as they expect (Cullen et al ., 1989; Kauffman, 1981; Klofas & Toch, 1982). Officers in this sample also perceive that inmate s in their facility are likely to blame inmate victims for rape. Paired t -tests reveal a significantly larger mean for officer belief in inmate tendency to blame compared to o fficers own levels of blaming. In other words, though officers have relatively low levels of blaming attitudes, they expect inmates in their facilities to blame inmate victims of rape. The findings for male ra pe myth items are similar; while officers are not particularly prone to accept male rape myths, they expect that inmates in their facility do adhere to these beliefs surrounding male rape. The results of this chapter are reve aling and contribute to research on these topics in a jail setting. Speci fically, the analyses in this chapter uniquely contribute to understand of officer perceptions of the attitudes of their fellow officers regarding counseling roles and corruption of authority. It also examines officer perceptions of inmate attitudes regarding victim blame and male rape myths; and compar es these perceptions to their own attitudes.

PAGE 195

Table 7-1. Perceptions of fellow officer professional orientation paired sample t -tests % agreement a Mean (SD) b Lower Upper t-value c Counseling roles Attitudes: rehabilitation programs should be left to mental health professionals Perceptions: rehabilitation programs should be left to mental health professionals 55 73 1.45 (0.50) 1.27 (0.45) .12 .24 5.94** Attitudes: counseling is a job for counselors, not correctional officers Perceptions: counseling is a job for counselors, not correctional officers 54 73 1.47 (0.50) 1.27 (0.44) .15 .26 7.59** Attitudes: if a correctional officer wants to do counseling, he should change jobs Perceptions: if a correctional officer wants to do counseling, he should change j obs 48 1.52 (.050) 1.34 (0.47) .13 .24 6.54** 66 Punitive orientation Attitudes: rehabilitation programs are a waste of time and money d Perceptions: rehabilitation programs are a waste of time and money d 44 68 1.44 (0.50) 1.68 (0.47) -.30 -.18 -7.92** Attitudes: improving jails for inmates makes them worse for officersd Perceptions: improving jails for inmates makes them worse for officers d 38 51 1.38 (0.49) 1.51 (0.50) -.18 -.08 -5.31** Attitudes: a military regime is the best way of running a jail d Perceptions: a military regime is the best way of running a jail d 71 1.70 (0.46) 1.77 (0.42) -.12 -.02 -3.04** 195 77 Corruption of authority Attitudes: you cant ever completely trust an inmated Perceptions: you cant ever completely trust an inmate d 88 95 1.87 (0.33) 1.95 (0.21) -.12 -.04 -4.07** Attitudes: a good principle is not to get close to inmatesd Perceptions: a good principle is not to get close to inmates d 93 95 1.93 (0.25) 1.95 (0.22) -.05 .01 -1.15 Attitudes: a personal relationship with an inmate invites corruption d Perceptions: a personal relationship with an inmate invites corruption d 94 95 1.94 (0.23) 1.96 (0.21) -.04 .01 -1.04 Attitudes: you must keep conversations with inmates short and businessliked Perceptions: you must keep conversations with inmates short and businesslike d 83 88 1.82 (0.38) 1.88 (0.32) -.10 -.01 -2.51* Attitudes: if an officer is lenient with inmates they will take advantage of him/ her d Perceptions: if an officer is lenient with inmates they will take advantage of him/ her d 89 1.91 (0.29) -.05 92 1.92 (0.27) .02 -.73

PAGE 196

Table 7-1. Continued % agreement a Mean (SD) b Lower Upper t-value c Social distance Attitudes: a correctional officer should work hard to earn trust from inmates Perceptions: a correctional officer should work hard to earn trust from inmates 57 34 1.43 (0.50) 1.66 (0.48) -.29 -.17 -7.77** Attitudes: its important for a correctional officer to have compassion Perceptions: its important for a correctional officer to have compassion 69 43 1.31 (0.46) 1.57 (0.50) -.31 -.20 -8.83** Attitudes: you get to like the inmates you supervise Perceptions: you get to like the inmates you supervise 34 24 1.66 (0.47) 1.76 (043) -.16 -.05 -3.67** Attitudes: sometimes a correctional officer should be an advocate for an inmate Perceptions: sometimes a correctional o fficer should be an advocate for an inmate 38 21 1.63 (0.48) 1.78 (0.41) -.21 -.11 -5.99** Attitudes: the way to get respect from inmates is to take an interest in them Perceptions: the way to get respect from inmates is to take an interest in them 29 1.71 (0.45) -.11 -.01 23 1.77 (0.42) -2.21* a % agreement indicates % who agree with the item (attitudes) and % who believe others agre e with the item (perceptions), b 1 = agree, 2 = disagree, c Mean difference tested through paired t -test, d Reverse-coded item, p <.05, **p <.01, *** p <.001 196

PAGE 197

197 Table 7-2. Perception of inma te attitudes paired sample t -tests % agreement a Mean (SD) b Lower Upper t-value c Victim blaming Attitudes: inmates who have previously consented get what they deserve if raped d Perceptions: inmates who have previously c onsented get what they deserve if raped d 18 64 1.18 (0.38) 1.64 (0.48) -.52 -.40 15.05** Attitudes: some inmates deserve to be raped because of the way they act d Perceptions: some inmates deserve to be raped because of the way they act d 10 61 1.10 (0.30) 1.61 (0.49) -.57 -.46 17.89** Attitudes: homosexual inmates get what they deserve if they are rapedd Perceptions: homosexual inmates get what they deserve if they are raped d 8 63 1.08 (0.27) 1.63 (0.48) -.60 -.49 18.52** Attitudes: inmates who take money/ cigarett es for sex get what they deserve if raped d Perceptions: inmates who take money/ cigarettes for sex get what they deserve if ra p ed d 16 67 1.16 (0.36) 1.67 (0.47) -.57 -.46 17.78** Attitudes: inmates who dress/ talk in femini ne ways get what they deserve if raped d Perceptions: inmates who dress or talk in feminine ways get what they deserve if ra p ed d 11 70 1.11 (0.32) 1.70 (0.46) -.64 -.53 21.38** Attitudes: most men raped by a man are somewhat to blame for not being careful d Perceptions: most men raped by a man are somewhat to blame for not being careful d 10 57 1.10 (0.30) 1.57 (0.50) -.53 -.41 15.75** Attitudes: most men who are raped by a man are to blame for not escaping or fighting d Perceptions: most men who are raped by a man are to blame for not escaping or fighting d 11 57 1.11 (0.31) 1.57 (0.50) -.52 -.41 16.54** Male rape myths Attitudes: most men who are raped by a man are very upset by the incident Perceptions: most men who are raped by a man are very upset by the incident 94 85 1.06 (0.24) 1.15 (0.36) -.13 -.05 -4.10** Attitudes: it is impossible for a man to rape a man dPerceptions: it is impossible for a man to rape a man d 7 21 1.08 (0.27) 1.20 (0.40) -.18 -.08 -5.15** Attitudes: most men who are raped by a ma n do not counseling after the incident d Perceptions: most men who are raped by a man do not counseling after the incident d 8 34 1.08 (0.27) 1.34 (0.47) -.31 -.21 -9.41** Attitudes: even a big, strong man can be raped by another man Perceptions: even a big, strong man can be raped by another man 88 71 1.12 (0.32) 1.30 (0.46) -.24 -.12 -5.95**a % agreement indicates % who agree with the item (attitudes) and % who believe inmate agree with the item (perceptions) b 1 = agree, 2 = disagree, c Mean difference tested through paired t -test, d Reverse-coded item, p <.05, **p <.01, *** p <.001

PAGE 198

CHAPTER 8 CONCLUSION AND DISCUSSION Introduction One of the premiere strengths of this study is that it explores a topic that has been ignored by social science research. The topic is one th at has obtained sufficient recent attention in the political arena and has garnered additional funding for prisons and jails nationwide. There have been no other recent in-depth stud ies of jail correctional officer perceptions of these topics using these analysis techniques. Specifically, no other studies have examined how jail correctional officers perceive inmate victims of sexual assault. This study al so uniquely examines the impact of belief in male rape myths on perceptions of inmate victims, significa nt differences between perceptions of the culpability and credibility of victims, officer definitions of rape, and their willingness to respond to rape in a jail setting. As long as sexual assault occurs in any context, there are implications for victims. While prior research has examined how those in law enforcement perceive this crime, very few have examined correctional officer attitudes regarding inmate victims of sexual assault or rape. Though the prevalence of sexual assault due to physical force or coercion is unknown, there is no dispute that it does occur in co rrectional institutions. Another contribution of this study is further examination of officer attitudes of in mates and their profession and the attitude s of their fellow officers. Lastly, this study offers a unique look at officer perceptions of inmate attitudes regarding inmate victims of sexual assault. Lessons The primary findings of this study will be pr esented according to the research questions outlined in Chapter 1 and reviewed in Chapter 3. 198

PAGE 199

Victim Blaming The first research question regarding percep tions of inmate vic tims of sexual assault explored in this study is: How do correctional officers perceive victims of sexual assault in jail? Specifically, do they attribute blame to victims of ra pe or sexual assault (victim-blame)? Overall, descriptive statistics reveal that this sample of officers has relatively low levels of victim blaming attitudes toward inmate victims of sexual assault and rape. Males in this sample are significantly more likely to blame victims than females. The t -tests reveal that this sample of officers is similar to samples elsewhere that as sign relative levels of blame based on various characteristics of and types of victims (Be ll et al., 1994; Calhoun et al., 1976; Eigenberg, 1989; Pugh, 1983; Schuller & Hastings, 2002). Howeve r, the way in which officers attribute culpability differs somewhat; sexual orientation is not as important in this context as it may be for blaming non-incarcerated vic tims of sexual assault. This finding is somewhat surprising because the sexual assault literature suggests that compared to heterosexual victims, homosexual victims are seen as more responsible for their victimization (Davies et al., 2006; Davies & Rogers, 2006; Ford et al., 1998; Mitchell et al 1999). Unlike studies of non-incarcerated victims, this study did not do a strict comparis on of homosexual victims versus heterosexual victims. In analyzing culpabi lity, homosexual inmates were compar ed to other types of inmate victims, and were never blamed more than ot her inmates were. This finding suggests that officers might expect homosexual inmates to be targeted for sexual assault in a correctional environment; and may therefore avoid blaming th em when they are targeted. This finding implies that the correctional culture (i.e., that vu lnerable inmates are targets) may matter more in assigning culpability than perceptions of victims or the culture of sexual assault. 199

PAGE 200

Comparing inmate victims reveal s that previously consenting to sex with other inmates or having exchanged sex for goods such as money or cigarettes makes inmates significantly more blameworthy in this study. Eigenberg (1989) also found that many of the respondents in her prison sample believed inmates who have previo usly consented to sex deserve it if raped by other inmates. Though some prior research (Howard, 1984a, 1984b) indicates that male rape victims are perceived as more bl ameworthy if they do not fight back, inmates in this sample who do not fight back are only rela tively blameworthy. Again, thes e findings may be explained by the culture of the institution and staff expectati ons of inmates. Officers may hold inmates more responsible for what they do rather than who th ey are while incarcerated. This may be true especially if inmates are active in the inmate cult ure and willingly engage in sex or the inmate economy. Likewise, they may be more hesitant to blame inmates who simply do not fight when confronted with sexual assault beca use they do not perceive failure to fight back as an instigation to their victimization. An interesting finding of this study that contrasts with other research is that sex does not maintain a significant association with victim blaming in the multivariate model, though prior studies indicate its importance in explaini ng blaming attitudes (B urczyk & Standing, 1989; Mitchell et al., 1999; White & Robinson Kurpi ous, 2002; Whatley & Riggio, 1993). The effect of male rape myths is no t surprising, as other studies report th at those who adhere to female rape myths are also more likely to blame female victim s for sexual assault (Frese et al., 2004; Jenkins & Dambrot, 1987; Kopper, 1996; Mason et al., 2004). What is interesting is that for this sample, male rape myth acceptance is more strongly associated with victim blaming attitudes than sex or attitudes toward homosexuality ar e in blaming inmate victims. The effect of attitudes toward homosexuality confirms that of prior studies where blaming homosexuals is more likely among 200

PAGE 201

those with less favorable att itudes toward them (Anderson, 2004; Burt & DeMello, 2002; White & Robinson Kurpius, 2002). Inmate Credibility The second research question about inmate victims of sexual assault that this study examines is: How do officers perceive the credibility of vi ctims of sexual assault in jail? Are all inmates equally credible as victims of sexual a ssault, or are there ce rtain inmates that are more likely to be believed as victims? According to Chapter 4, officers in this sample appear to give credibility to inmate victims of rape. However, Chapter 5 reveals that they consistently rank some inmates as more credible than others when reporting rape victimization. Muscular inmates, gang members, those who are drunk or high, or inmates who had previously consented to sex with other inmates are perceived as the least credible when reporting rape. Inma tes who delay reporting or have mental health problems are more believable than some inmates are, yet less believable than others are. Finally, young inmates are the most credible inmates when reporting rape for this sample. Homosexual inmates are also highly credible as victims when they report rape. That o fficers in this sample do not question reports from inmates who owe money is interesting, suggesting that these inmates may be routinely targeted for victimization. In comparing the results of the credibility anal ysis regarding percepti ons of inmates, these findings are somewhat consistent with those from other studies of both incarcerated and nonincarcerated victims. Female rape victims are also seen as less credible if they are drunk or stoned, delay reporting, have psycho logical problems, or have prev iously consented to sex with their accused (Brownmiller, 1975; Jordan, 2004; LaFree et al., 1985; Schuller & Stewart, 2000; Torrey, 1991). Eigenberg (1989) found that officer s in a prison setting were also likely to believe young inmates, inmates who are in debt and homosexual inmates; yet reluctant to 201

PAGE 202

believe muscular inmates and inmates who are ga ng members. This study reveals that inmates who have previously consented to sex with other inmates are the most blameworthy and the least credible as victims of rape. In contrast, officer s in this sample perceive homosexual inmates as the least blameworthy and the most credible. These findings are intere sting, given that prior research has indicated that compared to hete rosexual victims, homosexual victims are more culpable for sexual victimization (Davies et al ., 2006; Davies & Rogers, 2006; Ford et al., 1998; Mitchell at al., 1999). Correctional contexts, how ever, may be more important in explaining these results for homosexuals than the perceptions of victim sexual orientation. In other words, correctional officers, like inmates, may expect cert ain types of inmates to be targeted for sexual victimization, particularly homosexuals (John son, 2002; Nacci & Kane, 1984; Santos, 2004). These results also suggest that they do not auto matically assume the presence of consent in incidents involving homosexual inmates. The results of the multivariate model expl aining inmate credibility confirms the t -test results in Chapter 4 that indicate that males ar e less likely than females to believe inmates who report rape. The model also reveals that factors important in perceptions of inmate credibility differ somewhat across respondent sex. Experiencing job stress is significantly associated with believing inmate reports of rape for the full sample and male officer multivariate models. This result may be explained by officer percepti ons of dangerousness of the job and of the institutional environment (Castl e & Martin, 2006). For the full sample and female officers, victim blaming is the most important factor in explaining perceptions of inmate credibility. For males only, attitudes toward homosexuality is a significant factor rega rding perceptions of inmate credibility; males who have favorable atti tudes are more likely to perceive inmates as credible. Some researchers, in explaining differe nces in blaming attitudes about male victims of 202

PAGE 203

rape, especially homosexual victims, have stipulat ed that violations of sex role expectations regarding homosexuality attributed to blaming male victims may be especially salient for male respondents regarding male rape (Anderson, 200 4; Davies, 2004; Davi es, Pollard, & Archer, 2006; Herek, 1986; White & Robinson Kurpius, 2002). This may be one explanation why attitudes toward homosexuality was a factor for male (but not female) respondents in perceptions of inmate credibility. Also interesting regarding perceptions of inmate credibility is that while male rape myths are important in believing inmates for both female and males in the sample, the type of rape myth that matters differs across sex. Respondent sex differences in rape myth acceptance have generally analyzed only beliefs in rape myths about females (Quackenbush, 1989). The findings of this study support prior studies that find that perceptions of rape victims may be partially influenced by sex and subscription to rape myths (see Chapter 2). It is not surprising then that rape myths matter in explaining inmate believabi lity, as they also matter in perceptions of culpability. What is interesting, however, is th at different myths expl ain likelihood of believing inmate accounts of rape for male and female respondents. Perceptions of inmate credibility are less likely for males who adhere to the myth that men cannot rape other men and for females who adhere to the myth that men wh o are raped do not need counseling. The differing effects of type of male rape myth among male and female respondents may be a function of other variables thought to in fluence acceptance of rape myths, like attitudes about gender roles and expectati ons, and the nature of sexual relationships (Anderson, Cooper, & Okamura, 1997; Davies & Rogers, 2006). Specifica lly, perceptions of rape myths and credibility for male victims may be driven by gender-specific beliefs about the physical expectations for 203

PAGE 204

male behavior. It appears, the n, that whether male respondents in this sample believe reports of rape depends partially on whether they perceive rape between inmates as physically possible. Males in this sample who do perceive male rape as possible are more likely to believe reports of rape. Though perceptions of sex role expectations we re not tested, males may have more stringent and traditional expectations regarding physical male role expectations than females in the sample, believing that males who act outside these expectati ons are less credible as rape victims (Davies & McCartney, 2003; Herek, 1986). Some research suggests that adherence to traditional sex roles, including ideas about strength and masculinity, are endorsed more by males than by females (Davies & Rogers, 2006). Struckman-Johnson and StruckmanJohnson (1992) found that male subjects in thei r sample responded more dramatically than female subjects to the myth that men should be able to escape their attacker. For male respondents in this sample, viol ations of physical st rength expectations for men may preclude authenticity of male-on-male rape incidents; yet for female res pondents, violations of emotional expectations for men reduce percepti ons of victimization credibility. Female subjects who dismiss the myth that raped males do not need counseling are more likely to view inmate reports of ra pe as credible. In explaining leniency toward perpetrators who psychologically harmed male victim s, Schneider et al. (1994) stipul ated that male rape victims are perceived as less emotionally vulnerable than female victims are. One of the sex role expectations for men is that they are in cont rol of their emotions; therefore, reliance on counseling violates this expectation (Brannon & Juni, 1984; Hartley, 1959; Sawyer, 1970; Thomspon & Pleck, 1986). Females, because they may identify with rape victims more readily, specifically regarding emotional effects of rape, may be more likely to believe male inmate victims if they also believe males will respond emotionally to rape (Krulewitz, 1981; White & 204

PAGE 205

Robinson Kurpious, 2003). The effects of male rape myths across sex indicate that the manner in which beliefs about the culture of rape influen ce perceptions of the auth enticity of rape reports may depend on gendered differences in sex role expectations. Rape Definitions The third research question re garding sexual assault victimiz ation among inmates in jail examined in this study is: How do officers define rape in a jail atmosphere? Specifica lly, what are their attitudes regarding sexual consent af ter coercion or threats? According to Chapter 4, the sample has relatively inclusive attitudes about what constitutes rape. The paired sample t -tests reveal support for a prior study; officers in this sample are also rather consistent in defining situations that involve physical fo rce or violence threats as rape (Eigenberg, 2000a). This study uniquely contribute s to understanding of this issue by examining mean differences across scenarios to determine if officers are significantly more likely to define some situations over others as rape. Officers in this sample are significantly more likely to define a scenario as rape when it involves physical force or threats of violence. They are least likely to define a scenario as rape when it i nvolves quid pro quo situa tions where the inmate victim receives protection or goods prior to su bmitting to sex. Officers are less sure about defining as rape the situation where the inmate is ve rbally threatened with th e label of informant. Studies of non-incarcerated victims also suggest that definitions of rape often depend on context and cues about the behavior of the involve d parties (Amir, 1971; Campbell & Johnson, 1997; Estrich, 1987; Feldman-Summers & Palmer, 1980; Weis & Borges, 1975). The multivariate analysis of definitions of rape reveals that variables that were significant factors in a prior study of pr ison officers were not important here (Eigenberg, 2000a). Specifically, there was no significant effect of attitudes about victim blaming, professional 205

PAGE 206

orientation (such as counseling roles, punitive or ientation, corruption of authority, or social distance), or rape myth beliefs on officer defini tions of rape. Officers who perceive inmate victims as credible and older officers are significan tly more likely to define situations as rape. Other studies indicate that olde r officers have more positive eval uations about inmates compared to younger officers; although not measured, these ev aluations may help shape officer definitions in this sample (Cullen et al., 1989; Farkas, 1999; Griffin, 2002; Jurik, 198 5; Kifer et al., 2003; Klofas, 1986; Maahs and Pratt, 2001; Toch & Klofas, 1982; Teske & Williamson, 1979; Whitehead et al., 1987). The resu lts of the victim blaming analys is indicate that officers who do not believe inmate reports of rape are more likely to blame victims for their victimization. According to Chapter 6, these officers are also less likely to define situations as rape than respondents who believe inmate reports of rape. It is apparent that per ceptions of credibility, culpability, and rape definitions are c onnected for respondents in this sample. Willingness to Respond A fourth research question explored in this study regarding sexual assault victimization among inmates in jail asks: How do officers perceive their own willingness to respond to acts of sexual assault and rape in jail? Comparisons of the simple means for willingness to respond indices in Chapter 4 and the respective t -test results in Chapter 6, it appears that officers are most willing to respond to consensual sex and sexual assault by encour aging inmate reporting and officer prevention practices. They are least willi ng to talk to inmates about th ese incidents. They appear moderately willing to use proactive measures like cell assignments and protective custody to protect inmates from sexual assault, however. Th is finding is similar to that of prior research; 206

PAGE 207

officers are more supportive of proactive respons es than talking to inmates (Eigenberg, 1994, 2000b). Comparing responses to consen sual sex between inmates, o fficers are most comfortable doing everything they can to preven t these incidents than they are encouraging inmates to report them. They seem least comfortable talking to in mates about consensual se x. Officers also prefer to do everything they can to prevent and encour age inmate reporting of sexual assault to using proactive measures or talking to inmates about the risk s of sexual assault. They are supportive, however, of using cell assignments to combat sexual assault than in using protective custody or talking to inmates. Similar to incidents of cons ensual sex, officers are l east willing to indicate that they should talk to inmates about the risk of sexual assault. Overall, officers appear most comfortable doi ng everything they can to prevent consensual sexual acts or sexual assaults. They appear much less comfortable talking to inmates about these incidents, although they do indicate they are willing to encourage inma tes to report them. Comparing type of incident, officers are much more willing to respond to incidents of sexual assault than to incidences of consensual sex. This finding implies that they recognize the seriousness of assaultive over cons ensual sex and recognize their role in addressing the issue. In predicting willingness to respond to these incidents, demographic and work variables, attitudes about homosexuality, perceptions of inmate cred ibility, professional orientation items, and male rape myth items are all important; however, they vary in their effect on their willingness to respond. Encouraging inmate reporting and prevention Education, intolerance of homose xuality, perceptions of inmate s as credible, concern with corruption of authority, and disregard of rape my ths significantly predict willingness to address consensual sexual acts and sexua l assault by encouragi ng inmate reporting a nd officer prevention 207

PAGE 208

practices. These results are similar for the full sa mple and for officers without a college degree. Though the results for officers with a college degr ee differ somewhat, the coefficient comparison test reveals no significa nt difference in the effect of th ese variables on willingness between officers with or without a degree. Perceptions of inmate credibility in reports of rape is consistently important in predicting officer willingness to respond by encouraging inmates to report and officers to prevent consensual sex and sexual assault. This indicates that if officers believe that rape can happen in th eir institution, they are willing to act. The finding regarding corruption of authority is surprising because prior research suggests that other professional orientation items are mo re important in examining willingness to respond (Eigenberg, 1994). For this study, however, officers who perceive that it is possible for inmates to corrupt their authority indicate that they are more willing to encourage inmates to report consensual sex acts and sexual assault and to beli eve that officers should do everything they can to prevent these incidents. The results imply th at officers in this sample who believe inmate reports of rape and are concerned with inmate manipulation of their auth ority may also believe that ignoring these incidents can undermine the legitimacy of thei r authority and their role as well as threaten their professionalism (Bow ker, 1980; Wooden & Parker, 1982). This finding could also reflect affiliation for the aspect of the job dealing with enforcement of rules and control of inmate behavior. In other words, o fficers may perceive that one of their main job functions is to address these inci dents as rule violations, and may perceive getting close to or trusting inmates as a threat to their ability to perform this job function and to their professionalism. Similar to results in prior research, however, condemnation of homosexuality also predicts a willingness to respond in this manner, suggesting that these officers may believe 208

PAGE 209

that addressing the problem of sexual assaul t and consensual sex among inmates may also address homosexuality in jails (Eigenberg, 1994). Talking to inmates Experience, perceptions of inmate credibility and professional orientation items are factors that are more important when it comes to talking to inmates about these incide nts. In contrast to encouraging inmates to report and officer to prev ent, counseling roles and social distance are the professional orientation items that matter when it comes to talking to inmates. Officers with more experience, those who accept counseling roles, and those w ho prefer less social distance with inmates are significantly more willing to talk to inmates about consensual sex and about the risk of sexual assault. This finding suggests that officers who have spent more time around inmates and who embrace the human service aspe ct of correctional work are more open to discussing these sensitive topics with inmates. Perceptions of inmate credibility, counseling roles and social distance are also important; th ose who believe inmates who report rape, embrace counseling, and prefer less social distance w ith inmates are more willing to respond to consensual sex and sexual assa ult by talking to inmates. This finding supports Eigenbergs (1994) findings about the importance of counseling roles and social distance. Proactive measures Officers in this sample are quite willing to respond by using proactive measures to safeguard inmates from sexual assault (see Chapter 4). Officers who prefer less social distance in relationships with inmates, those who believe that inmate re ports of rape are credible, and those who are older are significantly more willing to respond with proactive measures. These findings regarding social distance are similar to previous research (E igenberg, 1994). Officers who agree that they should have compassion fo r and act as advocates for inmates are more willing to talk to inmates about ri sks of sexual assault and to use ja il administrative procedures to 209

PAGE 210

safeguard them. The results for the social distan ce variable indicates that officers concerned with compassion and advocacy for inmates are more willing to take proactive measures like talking with and safeguarding them (compared to simply encouraging inmate reporting and generally believing that officers should do what they can to prevent these acts). In explaining the three willingness to respond indices, age sometimes matters; older respondents are more likely to i ndicate that they are willing to use proactive measures to safeguard inmates from sexual assa ult. Older age is particularly important for officers with a college degree in predicting a willingness to respond by encouraging inmates to report and officers to prevent. More experience as a corr ectional officer is significantly correlated with willingness to respond by talking to inmates about consensual sex and the risks of sexual assault in the multivariate model. These results imply that officers with more life or correctional experience are more willing to respond to thes e incidences, and suggests support for research that indicates a link between age and positive orie ntations about inmates and the human service function of the correctional o fficer job (Cullen et al., 1989; Farkas, 1999; Griffin, 2002; Jurik, 1985; Kifer et al., 2003; Klofas, 1986; Maahs & Pratt, 2001; Toch & Klofas, 1982; Teske & Williamson, 1979; Whitehead et al., 1987). Professional orientation items are important, though there are differe nces regarding which item explains which willingness to respond index. Officers concerned wi th the corruption of their authority are more likely to respond by encouraging inmate reporting and officer prevention. Officers who accept counseling roles and prefer less social distance in their relationships with inmates are mo re willing to talk to inmates about consensual sex and the risk of sexual assault. Social distance is also a significant factor in explaining willingness to use proactive measures; officers who prefer less so cial distance are also more willing to respond 210

PAGE 211

proactively to sexual assault in jail by usi ng cell assignments and protective custody. A consistent finding across this analysis is that o fficers who perceive inmate reports of rape as credible are more likely to indicate willingne ss to respond by encouragi ng inmates to report and officers to prevent, by talking to inmates, and by using proactive measures to safeguard inmates. Perceptions of Fellow Officers Professional Orientation The next research questions re garding officer perceptions of the attitudes of their follow officers ask: Are correctional officers in jails in Florida accurate in their per ceptions of how other officers regard inmates and their professi on? How precisely do correctional officers identify their colleagues beliefs regardi ng counseling roles, punitive orientation, the corruption of authority, and so cial distance with inmates? This study offers a distinct contribution to the research about corre ctional officers by examining their attitudes and their perceptions of the attitudes of their fellow officers regarding two professional orientation item s: counseling roles and corruption of authority. The results suggest that correctional officers in this sample are plur alistically ignorant w ith regard to their perceptions of their fellow offi cer attitudes. Specifically, o fficers in the sample are more favorable toward counseling roles as part of thei r job function, they are less punitive, and they prefer less social distance between themselves and inmates than they give themselves credit for. They are also somewhat less concerned about the ability of inmates to corrupt their authority than they perceived themselves to be. These re sults are consistent with prior studies that find officers more accepting of rehabilitation, less custody oriented, and less aversive to inmates than they perceive (Cullen et al., 1989; Ka uffman, 1981; Klofas & Toch, 1982). Perceptions of Inmate Attitudes Finally, the last research que stion explores correctional officer perceptions of the attitudes of inmates, specifically: 211

PAGE 212

How do officer perceptions of victim blami ng and male rape myths compare to their perceptions of inmate att itudes regarding victim blam ing and male rape myths? In comparing the attitudes of offers to attit udes that officers perceive inmates to have, it appears that officers perceive th at inmate tendency to blame is significantly greater than their own tendency to blame. They also tend to percei ve that inmates in their facilities will have a greater tendency to acce pt male rape myths. Implications Practical Implications There are several implications of this study for the management of jail facilities and training of correctional officers. In correctional institutions, inmate reluctance to report sexual assault or rape can be attributed to fear of stigma, harassment, and victimization (Chonco, 1989; Lockwood, 1980; Weiss & Frirar, 1974). Victims of sexual assault also generally do not report for fear of being blamed or seen as lying (H odge & Cantor, 1998; Walker et al., 2005). The results of this study, however, reveal that officers are hesitant to blame victims and perceive their reports of rape as relatively credible. Percep tions of culpability and credibility, however, are connected; officers who do blame victims are also likely to question their reports of rape. If officers are hesitant to blame victims, one solution to address this issue would be to reduce the effects of the other inhibitors of inmate reporting. Specifically, it is essential for jail administrators to address further harassment a nd victimization of inmates who do report these incidents. Though jail administra tors are unlikely to influence the inmate culture and stigma surrounding victims, they can make response to sexua l assault a priority in their institutions. One way to incorporate this priority is to incr ease training of correctional officers to sensitize them to consent issues and myths surrounding the rape of males generally, and inmates in particular. For example, traini ng that focuses on consent can be helpful in clarifying that having 212

PAGE 213

previously submitted to sex or accepted protection in exchange for sex may not necessarily translate to full consent in a co rrectional environment. In add ition, facilities should talk to inmates about the risks of sexual assault upon intake, and use classification and housing procedures to identify and protect possible ta rgets (including homosexuals and those who have been previously assaulted) (McFarland et al., 2008). Jail administrators should ensure that their st aff are familiar with and understand protocol on how to deal with these incidents, as some research indicates that officers are not always familiar with these procedures (National Instit ute of Corrections & The Moss Group, Inc., 2006). Interestingly, a recent report of sexual violence in jails indicates that facilities with high incidents of sexual assault do not have adequate training about the PREA or the investigation, prevention, and reporting of sexual assa ult (McFarland et al., 2008). Offici als should also encourage officers to address what they perceive as consensual sexual acts between inmates as seriously as those they perceive as assaultive or forced. This study revealed that perceptions about previous consensual acts between inmates may shape o fficer attitudes about blame and credibility of reports. However, whether previous consensual act s are, in fact, consensual between inmates is not evident; the corr ectional culture may preclude c onsent (Bowker, 1980; Chonco, 1989; Tucker, 1992; Weiss & Friar, 1974; Wooden & Parker, 1982). A consistent finding of this study is that at titudes about inmate credibility matter for officers in this sample. Perceiving inmate reports of rape as credible is significantly correlated with whether officers avoid blaming, define inci dents as rape, and are willing to respond. Officers are more likely to perceive some inmates as more credible than other inmates, however. It is important to remember that this study does not indicate that reports of rape from all inmates are not taken seriously, only that some inmates appear more credible to officers. Facilities 213

PAGE 214

should continue to emphasize that all reports of rape should be taken seriously and investigated, regardless of victim characteristic s. Training about the myths su rrounding male rape (i.e., that men cannot rape other men) may al so assist in this endeavor. One of the ways in which officers are hesitant to address incidents of consensual sex and sexual assault with inmates is by talking to them about these issues. Though some (including this sample of officers) may argue that treatment staff may be be tter suited to car ry this burden, these types of staff may be more prevalent in prison than in jail settings (Stephan 2001, 2008). Therefore, it may be especially important for jail administrators who want officers to take on this responsibility to increase officer training in order to prepare them for it. Training that focuses on the importance of counseling, rehabilitation, a nd compassion for inmates may be useful, as officers who embrace these beliefs are more willing to talk to inmates about these incidents. This study reveals that officers in this sample are much more accepting of counseling roles and are less likely to require social distance between themselves and inmates than they believe. In other words, officers are relativ ely open to embracing counseling roles in their job and have compassion for inmates. Jail administrators should emphasize this reality to their staff and take advantage of the implications these attitudes have for professionalization. Their basic orientation and annual training, however, shoul d include training on PREA and the risks of sexual assault for inmates in these institutions. Some of the findings of this study indicate that age or experience are particularly important; specifically, older officers have more inclusive definitions of rape and those with more experience may be more willing to respond in certain ways. One of the practical applications for jail administrators regarding this finding is training of ne w jail officers. They 214

PAGE 215

may find it particularly helpful to pair offi cers who are older and have more correctional experience with younger new recrui ts during on the job training. Overall, correctional administrators should c ontinue to prioritize se xual assault and rape investigations to understand why th ese incidents occur. They shoul d also increase their focus on detecting, investigating, and prev enting consensual sex between inmates, an area relatively ignored by PREA. It is important to understand if these incidents really occur for the reasons the myths around them have traditionally suggested. In other words, the culture of correctional institutions suggests that inmates are target ed because they are seen as young, weak or vulnerable, feminine, or homosexual, but there is little empirical data that has supported these conclusions (Chonco, 1989; Fishman, 1934; Weiss & Friar, 1974; Wooden & Parker, 1982). The majority of studies about this topic have cons isted of interviews with inmates and staff, and while informative; may be based on rumor and myth more than empirical i nvestigations of actual incidents. Official responses to sexu al assault among non-incarcerate d samples can influence the reporting practices of victims (Field, 1978; Fe ldman-Summers & Palmer, 1980; Weis & Borges, 1975). If correctional staff are w illing to believe inmate reports of sexual assault and are willing to take actions to address its occurrence, inmates should be enc ouraged to report it. Though it is generally understood that not all instances of male rape are reported, controlling official response to male rape in correctional e nvironments may help address the problem of unreported rape. A recent report suggests that jails with high incident s of sexual assault fail to refer instances that do occur for outside prosecution. If the problem remains hidden or unaddressed when discovered, it will not be taken seriously by inmates or correctional staff. 215

PAGE 216

It is especially important to consider the jail co ntext in light of the findings of this research. Jail atmospheres are very different from pris on atmospheres regarding inmate admission and release, classification of inmates, resources, and staff. Jails ma y find it particularly difficult to address these issues given the s hort stays of some inmates and their limited housing resources. Prison administrators have more ability to use facility transfers to protect inmates; jail administrators are limited to the housing resource s they have available to separate vulnerable inmates from potential perpetrators. In addition, the importance of these issues in jail facilities may be diminished when inmates have relatively s hort stays, compared to longer prison stays. As discussed in Chapter 2, jail agencies that serve the dual roles of count y sheriff and county jail may find it particularly difficult to address staff responses to these issues. Theoretical Implications This study was not designed to test specific theories about attitudes; however, there are some general implications for theories of victim blaming and influences on correctional officer attitudes. One of the main theoretical approaches to blaming that has been explored primarily in literature on sexual assault is the just world perspective. Psychol ogist Melvin J. Lerner proposed a just world theory of victim blame (Lerner, 1970) by stipulati ng that people want to believe that others fates (both good and bad) are the result of their own actions, and that people get what they deserve (1970: 208). Some research supports Lerne rs contention that peoples perceptions of a just world are threatened when others receive harm or bad fates unjustly, and that people attempt to comp ensate and generalize about why undeserved harm befalls undeserving people. When belief in a just world is threatened, people will use tactics to adjust scripts and rationalizations to coincide with a just world (Correia & Vala, 2003; Crome & McCabe, 2001; Dalbert, 1999; Rubin & Pepla u, 1975; Whatley & Riggio, 1993). Lerners research further suggests that even when faced with an innocent victim, respondents will attempt 216

PAGE 217

to demean the victim rather than accept that the world may not be just. The rationale for this is simple as it has implications for the fate of ot hers: if we live in an unjust world, we can all succumb to fates that we do not deserve, no ma tter our good nature or hard work (Lerner, 1970; Lerner & Miller, 1978). In a ddition, Lerner & Miller (1978) s uggest that we become more concerned as a victims undeserved fate co mes closer to our wo rld and everyday life. If it appears that someone succumbs to a fate th at they do not deserve, people will typically find a way to establish desert through behavioral blaming (i.e., he is a good person, yet he behaved inappropriately) rather than characterological blami ng (i.e. he is just a bad person) (Janoff-Bulman, 1979, 1982). The findings regardi ng victim blaming in th is study suggest that officers may place more importance on actions or behaviors of inmates (e.g., previously engaging in sex or trading sex for goods) over character of inmates (e.g., having a homosexual sexual orientation) when assigning blame. Curre nt knowledge about percep tions of rape victim culpability indicates that victim characteristics ca n influence blaming attitudes and the just world theory of blaming suggests that blaming ma y serve as a self-protective mechanism. Though this study does not test these issues specifically, it rais es questions about correctional deprivation and importation theories Though traditionally applied to explanations of inmate adjustment and response to institutiona l environments, the theories have been explored in relation to correctional officer culture and beliefs. A handful of studies have attempted to explore whether demographic and background char acteristics of officer s provide the best explanation of their attitudes (i mportation), or whether their att itudes are malleable and adjusted by the institutional environment (deprivation) (Jurik & Halemba, 1984; Van Voorhis, Cullen, Link, & Wolfe, 1991). Though correctional officer attitudes are likely a combination of both importation (i.e., sex, age, pre-ex isting attitudes) and deprivation (i.e., char acteristics of the 217

PAGE 218

environment such as hyper-masculinity, etc.), it is difficult to isolate which theory offers the most explanation without longit udinal data (Jurik & Halemba, 1984; Van Voorhis et al., 1991). For example, cross-sectional data does not a llow determination of whether factors that are important in this study (such as acceptance of male rape myths and attitudes toward homosexuals) are imported or adopted from the co rrectional environment. These characteristics may be intrinsic in officers who apply for ja il work, or they may result from environmental characteristics, such as a jail facility do minated by male officers and male inmates. Limitations and Future Research The primary weakness of the analysis is that it focuses on one state and is therefore limited in generalizability. The data gathered is applicable to correctional officers in the Florida jails that were willing to participate in the researc h. It is possible that ja ils that were willing to participate in the research represent the more progressively administra ted jails and therefore results are not generalizable to othe r Florida jails. Similarly, the re sults may be an artifact of the types of officers who were willing to particip ate in the research by returning the completed survey. The questions in the survey also only ap ply to officer attitudes toward male inmates, which may further limit the generaliz ability of the results. While research about either female inmate sexuality or sex differences in sexuality among inmates is relatively rare, initial research suggests that the cultures differ (Castle, Hensley, & Tewksbury, 2002; Koscheski, Hensley, Wright, & Tewksbury, 2002; Struckman-Johnson & Struckman-Johnson, 2002; Tewksbury & West, 2000), Although most research sites for this study house both males and females and employ officers who supervise both sexes, it is beyond the scope of the present research to explore attitudes about both male and female jail inmates. Regardless, this study remains significant, as attitudes of jail correctional offi cers regarding these issues have not yet been examined. 218

PAGE 219

The data are drawn from self-reported attit udinal data, which depe nds on truthfulness of responses; and no objective measure was used to triangulate the research (i.e., observations of inmate and officer interactions, in-depth interviews with officer s, etc.). Though some of the material on the survey requested information on a ttitudes toward sensitive topics, other surveys of sensitive topics have shown to be valid, incl uding surveys of self-reported offending behavior (Junger-Tas & Marshall, 1999; Thornberry & Krohn, 2000). There is st ill the distinct possibility that correctional officers responded with what they perceive as s ocially desirable, rather than accurate, answers. However, some research sugg ests that respondents pr ovide accurate, rather than socially desirable information, if they feel that their responses will be kept confidential and anonymous (DeMaio, 1984; Schwarz, Strack, Hi ppler, & Bishop, 1991). In addition, selfadministered questionnaires are believed to decrea se socially desirable responses, compared to face-to-face interviews (Acquilino, 1994; Jones & Forrest, 1992; Schw arz et al., 1991; Turner et al., 1992). The low response rate indicates a possibil ity that correctional officers who actually returned the survey may differ from other correct ional officers on important characteristics, such as pronounced concern about inmates. Though the research netted a sufficient number of surveys for statistical purposes (i.e., between 100 and 1,000), the low response rate may bias the results of the research through lack of external validity and la ck of inferential power (Groves et al., 1992). Overall, the sample quality and ge neralizability are weak ened somewhat by the failure of most targets to participate (Bachman and Schutt, 2007). The cross-sectional design of the study limits its ability to establ ish causal order. Due to this lack of internal validity, it is impossible to determine whether officer attitude s are a result of thei r work experience, or 219

PAGE 220

whether they self-selected into this particular job based on preexisting characteristics. The results should be considered with these limitations in mind. Despite these limitations, the focus of this research is important as it examines a general topic that is gaining attention in the political and academic aren a, yet remains overlooked. There is very limited research on correctional officer attitudes toward the sexual victimization and behavior of inmates in general, and the research that does exist is focused on prison correctional officers. In addition, the resear ch on professional orientations of correctional officers has also been conducted primarily in state or federal prison institutions. This is unfortunate as jail facilities function as the primary entrance into the correctional system, and process a more dynamic population of inmates than prison facilities. This res earch will contribute to an understanding of jail correctional officers professional orientations, attitudes toward homosexuality, and attitudes toward the sexual assault and rape of those whom they supervise. This study reveals several avenues for futu re research. First, the study should be replicated in other areas and othe r types of jail fa cilities. It is impossible to understand whether these results are generalizable without extendin g its scope. Second, future research should consider a longitudinal approach to determine whether and how correctional officer perceptions change during their tenure. This type of study would contribute to a theoretical discussion of whether the importation or depriva tion approach best explains offi cer perceptions of inmates and their work environment. Though some prior stud ies have analyzed thes e questions in a prison environment, there is a paucity of research on the topi c regarding jail officers, and most studies are cross-sectional in design (Crouch & Al pert, 1982). A longitudi nal approach would contribute to our understanding of whether officer perceptions are imported into correctional environments (as suggested by Jurik & Halemba, 1984), or whether the nature of the work 220

PAGE 221

environment alters previously held attitudes (Jacobs & Kraft, 1978; Sykes, 1958). Future research would contribute to the correctional di scipline regardless of wh ether they are conducted in a prison or jail setting, as very few studies ha ve been conducted in this area. Another avenue for future studies is to survey inmates about their attitudes regarding blame, credibility, male rape myths, willingness to report, and tolerance of these incidents. It is likely that inmates have much different attitudes about these issues th an we understand as ma ny of the studies that interview inmates are either older or do not addre ss these particular issues (like male rape myths and blaming attitudes). Concluding Thoughts It appears that while some findings of this study mirror those found in studies of nonincarcerated sexual assault victims (i.e. the effect of rape myth beliefs on blaming), other results may be based more on the effects of the co rrectional culture. For example, while nonincarcerated homosexual victims are often seen as more blameworthy, officers in this sample attributed less blame to homosexual victims than to other types of inmates. This means that we cannot address the issue of sexual assault and rape in correctional institutions, including jails, without first addressing the culture of correctional institutions, including inmate and staff culture. Several studies show that correctional officers ar e progressively oriented toward rehabilitation goals of correctional institutions (Cullen et al., 1989; Eigenbe rg, 2000a; Farkas, 1999; Shamir & Drory, 1981; Toch & Klofas, 1982). Although officer s may be wary of taking on the role of a counselor or being lenient with inmates, they ma y not necessarily agree with being distant with inmates or uncompassionate (Farkas, 1999; Toch & Klofas, 1982). In order to address the issue of sexual assa ult in correctional institutions, we need to increase dialogue about and research in these areas. While PREA offe rs a good starting point, we need to ensure that the topic is not addressed only when there is a benefit of federal policy or 221

PAGE 222

222 fiscal resources, but that it becomes and remains a priority of correctional officials. The inmate experience is extremely important in light of current incarceration rates; so me of the highest this country has experienced (Blumste in & Beck, 1999; West & Sabol, 2008). If we are not willing to address these issues in correctional environments, then we have to accept that they may surface in our communities as most inmates ev entually come home (Petersilia, 2003; Travis, Soloman, & Waul, 2001).

PAGE 223

APPENDIX A SURVEY INSTRUMENT Circle the number next to the response that best describes your answer to each question. Please circle only one response to each question. There are no right or wrong answers. Remember all of your responses are anonymous. 1. What is your sex? (circle one) 01) Male 02) Female 2. Are you Spanish, Hispanic, or Latino ? (circle one) 01) No 02) Yes 3. What is your race? (circle all that apply) 01) White 04) Biracial 02) Black 05) Other: Please specify 03) Asian, Pacific Islander 4. What is your marital status? (circle one) 01) Married 04)Separated 02) Widowed 05)Never Married 03) Divorced 5. What is the highest degree or level of school you completed? (circle one) 01) Less than a high-school education 05) Associate degree (for example: AA, AS) 02) High school graduatehigh school 06) Bachelors degree (for example: BA, AB, 03) Some college credit, but less than 1 year 07) Masters degree (for example: MA, MS, 04)1 or more years of college, no degree 08) Doctorate or Professional degree (for 6. What shift do you typically work? (circle one) 01) Day shift (morning to afternoon) 02) Evening shift (afternoon to late evening or midnight) 03) Night shift (late evening or midnight to early morning) 7. How old are you? (circle the range that contains your age at your most recent birthday) 01) 18-25 06) 46-50 02) 26-30 07) 51-55 03) 31-35 08) 56-60 04) 36-40 09) Over 60 05) 41-45 8. Have any of your family or friends ever been incarcerated? (circle one) 01) Yes 02) No 9. What political party do you belong to? (circle one) 01) Democratic 03) Republican 02) Independent 04) Other: Please specify 10. How would describe your political orientation? (circle one) 01) Extremely liberal 05) Conservative 02) Very liberal 06) Very conservative 03) Liberal 07) Extremely conservative 04) Middle of the road 223

PAGE 224

11. On a scale of 1 to 10, how religious do you consider yourself (1 being not at all religious and 10 being extremely religious)? (write the number in the space at the ri g ht ) ___________ 12. On a scale of 1 to 10, how satisfied are you with your job (1 being not at all satisfied and 10 being extremely satisfied)? (write the number in the space at the ri g ht ) ___________ 13. On a scale of 1 to 10, how much stress do you experience on the job (1 being no stress and 10 being extreme stress)? (write the number in the space at the right) ___________ 14. How long have you been working as a correctional officer (at any facility)? (estimate to the nearest year and write the number in the space at the right) ___________ 15. About how many correctional officers work at your jail facility? (write the number in the space at the right) ___________ 16. About how many inmates are currently housed at your jail facility? (write the number in the space at the right) ___________ 17. Do you consider the facility in which you work to be located in an rural, urban, or suburban area? 01) Rural (it is in the open country or in a town of less than 2,500 people) 02) Urban (it is in an area with a total population of 50,000 or more) 03) Suburban (it is somewhere between rural and urban) Please indicate whether YOU agree or disagree with the following statements. Remember all of your responses are anonymous. Circle only one response for each question. Strongly Agree Agree Not Sure, Probably Agree Not Sure, Probably Disagree Disagree Strongly Disagree 18. Rehabilitating a criminal is just as important as making a criminal pay for his or her crime. 1 2 3 4 5 6 19. The only effective and humane cure to the crime problem in America is to make a strong effort to rehabilitate offenders. 1 2 3 4 5 6 20. The only way to reduce crime in our society is to punish criminals, not try to rehabilitate them. 1 2 3 4 5 6 21. We should stop viewing criminals as victims of society who deserve to be rehabilitated and start paying more attention to the victims of these criminals. 1 2 3 4 5 6 22. The rehabilitation of prisoners has proven to be a failure. 1 2 3 4 5 6 23. The rehabilitation of adult criminals just does not work. 1 2 3 4 5 6 224

PAGE 225

Strongly Agree Agree Not Sure, Probably Agree Not Sure, Probably Disagree Disagree Strongly Disagree 24. I would support expanding the rehabilitation programs with criminals that are now being undertaken in our prisons. 1 2 3 4 5 6 25. One of the reasons why rehabilitation programs often fail with prisoners is because they are under-funded. If enough money were available to hire trained personnel and to administer treatment properly, then these programs would work. 1 2 3 4 5 6 26. All rehabilitation programs have done is to allow criminals who deserve to be punished to get off easily. 1 2 3 4 5 6 27. You cant ever completely trust an inmate. 1 2 3 4 5 6 28. A good principle is not to get close to inmates. 1 2 3 4 5 6 29. A personal relationship with an inmate invites corruption. 1 2 3 4 5 6 30. You must keep conversations with inmates short and businesslike. 1 2 3 4 5 6 31. If an officer is lenient with inmates they will take advantage of him/ her. 1 2 3 4 5 6 32. A correctional officer should work hard to earn trust from inmates. 1 2 3 4 5 6 33. Its important for a correctional officer to have compassion. 1 2 3 4 5 6 34. You get to like the inmates you supervise. 1 2 3 4 5 6 35. Sometimes a correctional officer should be an advocate for an inmate 1 2 3 4 5 6 36. The way to get respect from inmates is to take an interest in them. 1 2 3 4 5 6 37. Rehabilitation programs should be left to mental health professionals. 1 2 3 4 5 6 38. Counseling is a job for counselors, not correctional officers. 1 2 3 4 5 6 39. If a correctional officer wants to do counseling, he should change jobs. 1 2 3 4 5 6 40. Rehabilitation programs are a waste of time and money. 1 2 3 4 5 6 41. There would be less crime if jails were more comfortable. 1 2 3 4 5 6 225

PAGE 226

Strongly Agree Agree Not Sure Probably Agree Not Sure, Probably Disagree Disagree Strongly Disagree 42. Improving jails for inmates makes them worse for officers. 1 2 3 4 5 6 43. A military regime is the best way of running a jail. 1 2 3 4 5 6 Thinking about these same questions, guess how other officers in your facility would answer each question. Tell me if you think all or most other officers would agree or disagree with the following statements. Circle only one response for each question. Almost All Agree [over 80% of the officers would agree] Most Agree [more than half but fewer than 80% would agree] Most Disagree [more than half but fewer than 80% would disagree] Almost All Disagree [80% or more would disagree] 44. You cant ever completely trust an inmate. 1 2 3 4 45. A good principle is not to get close to inmates. 1 2 3 4 46. A personal relationship with an inmate invites corruption. 1 2 3 4 47. You must keep conversations with inmates short and businesslike. 1 2 3 4 48. If an officer is lenient with inmates they will take advantage of him/ her. 1 2 3 4 49. A correctional officer should work hard to earn trust from inmates. 1 2 3 4 50. Its important for a correctional officer to have compassion. 1 2 3 4 51. You get to like the inmates you supervise. 1 2 3 4 52. Sometimes a correctional officer should be an advocate for an inmate. 1 2 3 4 53. The way to get respect from inmates is to take an interest in them. 1 2 3 4 54. Rehabilitation programs should be left to mental health professionals. 1 2 3 4 55. Counseling is a job for counselors, not correctional officers. 1 2 3 4 56. If a correctional officer wants to do counseling, he should change jobs. 1 2 3 4 57. Rehabilitation programs are a waste of time and money. 1 2 3 4 58. There would be less crime if jails were more comfortable. 1 2 3 4 226

PAGE 227

Almost All Agree [over 80% of the officers would agree] Most Agree [more than half but fewer than 80% would agree] Most Disagree [more than half but fewer than 80% would disagree] Almost All Disagree [80% or more would disagree] 59. Improving jails for inmates makes them worse for officers. 1 2 3 4 60. A military regime is the best way of running a jail. 1 2 3 4 Please indicate whether YOU agree or disagree with the following statements. REMEMBER ALL OF YOUR RESPONSES ARE ANONYMOUS. Circle only one response for each question. Strongly Agree Agree N ot Sure, P robably A gree Not Sure, Probably Disagree D isagree Strongly Disagree 61. Inmates who have previously consented to participate in sexual acts in jail get what they deserve if they are raped by other inmates. 1 2 3 4 5 6 62. Some inmates deserve to be raped in jail because of the way they act. 1 2 3 4 5 6 63. Homosexual inmates get what they deserve if they are raped in jail. 1 2 3 4 5 6 64. Inmates who take money or cigarettes in exchange for consensual sexual acts get what they deserve if they are raped by other inmates. 1 2 3 4 5 6 65. Inmates who dress or talk in feminine ways get what they deserve if they are raped in jail. 1 2 3 4 5 6 66. Most men who are raped by a man are somewhat to blame for not being more careful. 1 2 3 4 5 6 67. Most men who are raped by a man are to blame for not escaping or fighting off the man. 1 2 3 4 5 6 68. Most men who are raped by a man are very upset by the incident. 1 2 3 4 5 6 69. It is impossible for a man to rape a man. 1 2 3 4 5 6 70. Most men who are raped by a man do not need counseling after the incident. 1 2 3 4 5 6 71. Even a big, strong man can be raped by another man. 1 2 3 4 5 6 227

PAGE 228

Strongly Agree Agree N ot Sure, P robably A gree Not Sure, Probably Disagree D isagree Strongly Disagree 72. Sometimes it is hard to tell whether inmates are being forced to participate in sexual acts or if they are willing partners in consensual sexual acts. 1 2 3 4 5 6 73. Inmates who play the passive role (insertee) during intercourse are homosexuals. 1 2 3 4 5 6 74. Inmates who play the active role (inserter) during intercourse are homosexuals. 1 2 3 4 5 6 75. Under certain circumstances, heterosexual men may choose to participate in consensual sexual acts with other men. 1 2 3 4 5 6 76. Male homosexual couples should be allowed to adopt children the same as heterosexual couples 1 2 3 4 5 6 77. I would not be too upset if I learned that my son was a homosexual. 1 2 3 4 5 6 78. Homosexual behavior between two men is just plain wrong. 1 2 3 4 5 6 79. Male homosexuality is merely a different kind of lifestyle that should not be condemned. 1 2 3 4 5 6 80. Just as in other species, male homosexuality is a natural expression of sexuality in human men. 1 2 3 4 5 6 81. I think male homosexuals are disgusting. 1 2 3 4 5 6 82. Male homosexuals should not be allowed to teach school. 1 2 3 4 5 6 83. If a man has homosexual feelings, he should do everything he can to overcome them. 1 2 3 4 5 6 84. Male homosexuality is a perversion. 1 2 3 4 5 6 85. The idea of male homosexual marriages seems ridiculous to me. 1 2 3 4 5 6 228

PAGE 229

Thinking about these SAME questions, guess how INMATES in your facility would answer each question. Tell me if you think all or most inmates would agree or disagree with the following statements. Circle only one response for each question. Almost All Agree [over 80% of the inmates would agree] Most Agree [more than half but fewer than 80% would agree] [more than half Most Disagree b ut fewer than 80% would disagree] Almost All Disagree [80% or more would disagree] 86. Inmates who have previously consented to participate in sexual acts in jail get what they deserve if they are raped by other inmates. 1 2 3 4 87. Some inmates deserve to be raped in jail because of the way they act. 1 2 3 4 88. Homosexual inmates get what they deserve if they are raped in jail. 1 2 3 4 89. Inmates who take money or cigarettes in exchange for consensual sexual acts get what they deserve if they are raped by other inmates. 1 2 3 4 90. Inmates who dress or talk in feminine ways get what they deserve if they are raped in jail. 1 2 3 4 91. Most men who are raped by a man are somewhat to blame for not being more careful. 1 2 3 4 92. Most men who are raped by a man are to blame for not escaping or fighting off the man. 1 2 3 4 93. Most men who are raped by a man are very upset by the incident. 1 2 3 4 94. It is impossible for a man to rape a man 1 2 3 4 95. Most men who are raped by a man do not need counseling after the incident 1 2 3 4 96. Even a big, strong man can be raped by another man 1 2 3 4 Please indicate whether YOU agree or disagree with the following statements. REMEMBER ALL OF YOUR RESPONSES ARE ANONYMOUS. Circle only one response for each question. Strongly Agree Agree Not Sure, Probably Agree N ot Sure, P robably D isagree Disagree S trongly D isagree 97. Jail officers should encourage inmates to report consensual sexual acts that occur in jail. 1 2 3 4 5 6 229

PAGE 230

Strongly Agree Agree Not Sure, Probably Agree N ot Sure, P robably D isagree Disagree S trongly D isagree 98. Jail officers should encourage inmates to report sexual assaults that occur in jail. 1 2 3 4 5 6 99. Jail officers should do everything they can to prevent consensual sexual acts in jail. 1 2 3 4 5 6 100. Jail officers should do everything they can to prevent sexual assaults in jail. 1 2 3 4 5 6 101. Jail officers should talk to inmates about consensual sexual acts to discourage these activities. 1 2 3 4 5 6 102. Jail officers should talk to inmates about the risk of sexual assault in jail. 1 2 3 4 5 6 103. Jail officers should use cell assignments to safe-guard inmates from sexual assault. 1 2 3 4 5 6 104. Jail officers should refer inmates to protective custody to safe-guard them from sexual assault. 1 2 3 4 5 6 105. Inmate Jones physically overpowers inmate Smith. Smith has sex with Jones. Smith has been raped. 1 2 3 4 5 6 106. Inmate Jones threatens to tell other inmates that inmate Smith is an informant unless inmate Smith engages in sexual acts. Smith has sex with Jones. Smith has been raped. 1 2 3 4 5 6 107. Inmate Jones tells inmate Smith he will kill him unless Smith has sex with Jones. Smith has sex with Jones. Smith has been raped. 1 2 3 4 5 6 108. Inmate Smith is an informant. Inmates Jones provides protection for Smith but demands that Smith participate in sexual acts. Smith has sex with Jones. Smith has been raped. 1 2 3 4 5 6 230

PAGE 231

231 Strongly Agree Agree Not Sure, Probably Agree N ot Sure, P robably D isagree Disagree S trongly D isagree 109. Inmate Jones loans inmate Smith money or some goods. Smith cannot pay Jones back. Jones tells Smith that he can participate in sexual acts to pay off his debt or inmate Jones will beat inmate Smith severely. Smith has sex with Jones. Smith has been raped. 1 2 3 4 5 6 0. In jails in this state, about what percentage of inmates do you think have been pressured or forced to have sexual contact against their will by other inmates? (circle one answer that represents your best guess) 0% 1% 5% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% An inmate tells you he was raped. How likely are you to believe him if he: (circle one response for each question) Always Generally Sometimes Rarely Never 111. Is a muscular inmate 1 2 3 4 5 112. Is a homosexual inmate 1 2 3 4 5 113. Is a young inmate 1 2 3 4 5 114. Is a gang member 1 2 3 4 5 115. Is an inmate who owes money 1 2 3 4 5 116. Is drunk or high on drugs 1 2 3 4 5 117. Delayed reporting the incident 1 2 3 4 5 118. Has previously consented to sex with other inmates 1 2 3 4 5 119. Has mental health problems 1 2 3 4 5 Please feel free to use the space below and the back of the survey to elaborate further on any of your responses in the survey. Please do not incl ude your name or other information in this section that could be used to identify you. This is the end of the survey. Thank you for ta king the time to participate in this survey!

PAGE 232

LIST OF REFERENCES Abram, K.M. & Teplin, L.A., (1991). Co-occu rring disorders among mentally ill jail detainees: Implications for public policy. American Psychologist 46, 1036-1045. Acquilino, W.S. (1994). Interview mode effects in surveys of drug and al cohol use: A field Experiment. Public Opinion Quarterly 58 ,210. Agresti, A., & Finlay, B. (1997). Statistical methods for the social sciences (3rd ed.). Upper Saddle River, NJ: Prentice Hall. Allison, P.D. (2002). Missing data. Sage University Papers Series on Quantitative Application in the Social Sciences, 07-136. Thousand Oaks, CA: Sage. Alpert, G.P. & Crouch, B.M. (1991). Cross-gender supervision, personal privacy, and institutional security: Perceptions of jail inmates and staff. Criminal Justice and Behavior, 18: 304-317. Amir, M. (1971). Patterns in forcible rape. Chicago: University of Chicago Press. Anderson, I. (1999). Characterological & beha vioral blame in conversations about female & male rape. Journal of Language & Social Psychology, 18, 377-394. Anderson, I. (2004). Explaining nega tive rape victim perception: Ho mophobia and the male rape victim. Current Research in Social Psychology, 10, 43-57. Anderson, K.B., Cooper, H., & Okamura, L. (1997) Individual differences and attitudes toward rape: A meta-analytic review. Personality and Social Psychology Bulletin 23, 295-315. Applegate, B.K. & Sitren, A.H. (2008) The jail a nd the community: Comparing jails in rural and urban contexts. The Prison Journal, 88 252-269. Arthur, J.A. (1994). Correctional ideo logy of Black correctional officers. Federal Probation, 58 57-66. Bachman, R. & Schutt, R.K. (2007). The practice of research in criminology and criminal justice (3rd ed.). Los Angeles: Sage. Barnett, M.A., Quackenbush, S.W., Sinisi, C.S., Wegman, C.M., & Otney, K.L. (1992). Factors affecting reactions to a rape victim. The Journal of Psychology, 126 609-620. Beck, A.J. (2002). Jail population growth: Nationa l trends and predictors of future growth. American Jails, 16, 9-14. 232

PAGE 233

Beck, A.J. & Harrison, P.M. (2008). Sexual victimization in local jails reported by inmates, 2007. Washington, DC: US Department of Justice, Office of Justice Pr ograms, Bureau of Justice Statistics. Retr ieved March 18, 2009 from http://ojp.usdoj.gov/bj s/pub/pdf/svljri07.pdf Beck, A.J., Harrison, P.M., & Adams, D.B. (2007). Sexual violence reporte d by correctional authorities, 2006 Washington, DC: US Department of Justice, Office of Justice Programs, Bureau of Justice Statis tics. Retrieved December 8, 2008 from http://www.ojp.usdoj.gov/ bjs/pub/pdf/svrca06.pdf Beck, A.J., & Hughes, T.A. (2005). Sexual violence reported by correctional authorities, 2004 Washington, DC: US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. Retrie ved September 10, 2007 from http://www.ojp.usdoj.gov/ bjs/pub/pdf/svrca04.pdf Bell, S.T., Kuriloff, P.J., & Lottes, I. (1994). Understanding attributions of blame in stranger rape and date rape situations: An examinati on of gender, race, iden tification, and students social perceptions of rape victims. Journal of Applied Social Psychology, 24, 1719-1734. Blumstein, A. & Beck, A. (1999). Population grow th in U.S. prisons, 1980-1996. In M.Tonry & J. Petersilia (Eds.), Crime and Justice: A Review of Research, 26, 17-61. Bottoms, B.L., & Goodman, G.S. (1 994). Perceptions of childrens cr edibility in sexual assault cases. Journal of Applied Social Psychology, 24, 702-732. Bowker, L. (1980). Prison victimization. New York: Elsevier. Bowker, L. (1982). Victimizers and victims in Am erican correctional institutions. In R. Johnson & H. Toch (Eds.), The pains of imprisonment (pp. 63-76). Beverly Hills: Sage. Brannon, R., & Juni, S. (1984). A scale for measuring attitudes about masculinity. Psychological Documents, 14, 6-7. Bridges, J.S. (1991). Perceptions of date and stra nger rape: A difference in sex role expectations and rape supportive beliefs. Sex Roles, 24 291-307. Brownmiller, S. (1975). Against our will: Men, women, and rape New York: Simon & Schuster. Burczyk, K., & Standing, L. (1989). Attitudes towards rape victims: Effects of victim status, sex of victim, and sex of rater. Social Behavior and Personality, 17, 1-8. Bureau of Justice Statistics (2008). Death in custody statistical tables Washington, DC: US Department of Justice, Office of Justice Programs. Retrieved October 13, 2008 from http://www.ojp.usdoj.gov/bjs/dcrp/tables/jailstab2.htm 233

PAGE 234

Burt, M. (1980). Cultural myths and supports for rape. Journal of Personality and Social Psychology, 38, 217-230. Burt, D.L., & DeMello, L.R. (2002). Attribution of rape blame as a function of victim gender and sexuality, and perceived si milarity to the victim. The Journal of Homosexuality, 43, 3957. Calhoun, L.G., Cann, A., Selby, J.W., & Magee, D.L. (1981). Victim emotional response: Effects on social reactio n to victims of rape. British Journal of Social Psychology, 20, 1721. Calhoun, L.G., Selby, J.W., & Warring, L.J. (1976). Soci al perception of the victims causal role in rape: An exploratory exam ination of four factors. Human Relations, 29 517-526. Campbell, R. & Johnson, C.R. (1997). Police officers perceptions of rape: Is there consistency between state law and individual beliefs? Journal of Interpersonal Violence, 12 255-274. Cann, A., Calhoun, L.G., & Selby, J.W. (1979). Attrib uting responsibility to the victim of rape: Influence of information rega rding past sexua l experience. Human Relations, 32, 57-67. Castle, T., Hensley, C., & Tewksbury, R. (2002). Argot roles and prison se xual hierarchy. In C. Hensley (Ed.), Prison Sex: Practice & Policy (pp. 13-26). Boulder: Lynne Rienner. Castle, T.L. & Martin, J.S. (2006). Occupational hazard: Predictors of stress among jail correctional officers. American Journal of Criminal Justice, 31 65-80. Chapleau, K.M., Oswald, D.L., & Russell, B.L. ( 2008). Male rape myths: The role of gender, violence, and sexism. Journal of Interpersonal Violence, 23 600-615. Chonco, N. (1989). Sexual assaults am ong male inmates: A descriptive study. The Prison Journal, 69, 72-82. Clear, T.R., Cole, G.F., & Reisig, M.D. (2006). American corrections (7th ed.). Belmont, CA: Thompson Wadsworth. Clogg, C.C., Petkova, E., & Haritou, A. (1995). St atistical methods for comparing regression coefficients between models. American Journal of Sociology 100 1261-1293. Corlew, K.R. (2006). Congress attempts to shine a light on a dark problem: An in-depth look at the Rape Elimination Act of 2003. American Journal of Criminal Law, 33 157-190. Correia, I. & Vala, J. (2003). When will a victim be secondarily victimiz ed? The effect of observers belief in a just world, victim s innocence and persistence of suffering. Social Justice Research, 16 379. 234

PAGE 235

Cotton, D. J., & Groth, A. N. (1982). In mate rape: Prevention and intervention. Journal of Prison and Jail Health, 2, 47-57. Crome, S.A. & McCabe, M.P. (2001). Adult rape sc ripting within a victim ological perspective. Aggression and Violent Behavior 6, 395-413. Crouch, B.M & Alpert, G.P. (1982). Sex and occ upational socialization among prison guards: A longitudinal study. Criminal Justice and Behavior, 9, 159-176. Cullen, F.T., Latessa, E.J., Burton, V.S., & Lombar do, L.X. (1993). The correctional orientation of prison wardens: Is the reha bilitative ideal supported? Criminology, 31, 69-92. Cullen, F.T., Lutze, F.E., Link, B.G., & Wolfe, N.T. (1989). The correctio nal orientation of prison guards: Do officer s support rehabilitation? Federal Probation, 53 33-42. Dalbert, C. (1999) The world is more just for me th an generally: About the personal belief in a just world scale's validity. Social Justice Research, 12, 79. Davies, M. (2002). Male sexual assault victims: A selective re view of the literature and implications for support services. Aggression and Violent Behavior 7, 203-214. Davies, M. (2004). Correlates of negative attitudes toward gay men: Sexism, male role norms, and male sexuality The Journal of Sex Research, 41, 259-266. Davies, M., & McCartney, S. (2003) Effects of gender and sexuali ty on judgments of victim blame and rape myth acceptance in a depicted male rape. Journal of Community & Applied Psychology, 13 391-398. Davies, M., Pollard, P., & Archer, J. (2006). Effects of perpetrato r gender and victim sexuality on blame toward male victims of sexual assault. The Journal of Social Psychology, 146 275-291. Davies, M., & Rogers, P. (2006). Perceptions of male victims in depict ed sexual assaults: A review of the literature. Aggression and Violent Behavior, 11 367-377. Deitz, S.R., & Byrnes, L.E. (1981). Attribution of responsib ility for sexual assault: The influence of observer empathy and defendant occupation and attractiveness. Journal of Psychology 108, 17-29. DeMaio, T.J. (1984). Social desirability and surv ey measurement: A review. In C.F. Turner, & E. Martin. (Eds.), Surveying subjective phenomena vol. 2 (pp. 257-281). New York: Russell-Sage. Dempster, A.P., Laird, N.M., & Rubin, D.B. (1977) Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, 39 1-38. 235

PAGE 236

Dempster, A.P., & Rubin, D.B. (1983). Introduction. In W.G. Madow, H. Nisselson, & I. Olkin (Eds.) Incomplete data in sample surveys (Vol. 1). New York: Academic Press. Dennehy, K.M. & Natel, K.A. (2006). Improving pris on safety: Breaking the code of silence. Washington University Jour nal of Law and Policy, 22, 175-186. Dillman, D. (2000). Mail and internet surveys: The tailored method (2nd ed.). New York: John Wilery & Sons. Dillman, F., Dolsen, D.E., & Machlis, G.E. (1995) Increasing response to personally-delivered mail-back questionnaires. Journal of Official Statistics, 11 129-139. Doerner, W.G., & Lab, S.P. (2008). Victimology (5th ed.). Newark: Anderson. Doherty, K., & Anderson, I. ( 2004). Making sense of male rape : Constructions of gender, sexuality and experience of rape victims. Journal of Community & Applied Social Psychology, 14 85-103. Duffee, D. (1974). The correction officer subculture and organizational change. Journal of Research in Crime and Delinquency, 11, 155-172. Dumond, R.W. (2000). Inmate Sexual Assault: The plague that persists. The Prison Journal, 80, 407-414. Dumond, R.W. & Dumond, D.A. (2002). The treatment of sexual assault victims. In C. Hensley (Ed.). Prison sex: Practice and policy (pp. 67-87.). Boulder, CO: Lynne Rienner. Dynes, W. & Donaldson, S. (1992). Homosexuality & government, politics, and prisons New York: Garland. Eigenberg, H. (1989). Male rape: An empirical examination of co rrectional officers attitudes toward rape in prison. The Prison Journal, 68, 39-56. Eigenberg, H. (1994). Rape in male prisons: Ex amining the relationship between correctional officers attitudes toward male rape & their willingness to respond to acts of rape. In M. Braswell, R. Montgomery & L. Lombardo (Eds.), Prison violence (2nd ed.). (pp 145165). New York: Anderson. Eigenberg, H. (2000a). Correct ional officers definitions of rape in male prisons. Journal of Criminal Justice, 28, 435-449. Eigenberg, H. (2000b). Correctiona l officers & their perceptions of homosexuality, rape, & prostitution in male prisons. The Prison Journal, 80 415-433. Estrich, S. (1987). Real rape Cambridge, MA: Harvard University Press. 236

PAGE 237

Farkas, M.A. (1999). Correctional officer attitudes toward inmates and working with inmates in a get tough era. Journal of Criminal Justice 27, 495-506. Farkas, M.A. (2001). Correctional officers: What factors influence work attitudes? Corrections Management Quarterly 5, 20-26. Farmer v. Brennan, 511 U.S. 825 (1994). Feild, H.S. (1978). Attitudes toward rape: A compar ative analysis of poli ce, rapists, crisis counselors, and citizens. Journal of Personality and Social Psychology, 36, 156-179. Feldman-Summers, S., & Lindner, K. (1976). Perceptions of victims and defendants in criminal assault cases. Criminal Justice and Behavior, 3, 135-150. Feldman-Summers, S. & Palmer, G.C. (1980). Rape as viewed by judges, pr osecutors, and police officers. Criminal Justice and Behavior, 7, 19-40. Flannery, J.P. (1980). Prison co rruption: A mockery of justice, In I.P. Robbins (Ed.) Prisonersrights sourcebook: Theory, litigation, practice (Vol. 2). New York: Clark Boardman. Fishman, J. (1934). Sex in prison: Revealing sex c onditions in American prisons. New York: National Library. Fleisher, M. (2005). The culture of prison sexuality & rape Report submitted to the National Institute of Justice. Washington D.C.: Department of Justice. Foley, L.A., Evancic, C., Karnick, K., King, J., & Pa rks, A. (1995). Date rape: Effects of race on assailant and victim and gende r of subjects on perceptions. Journal of Black Psychology 21, 6-18. Ford, T.M., Liwag-McLamb, M.G., & Foley, L.A. (1998). Perceptions of ra pe based on sex and sexual orientation of victim. Journal of Social Behavior and Personality 13, 253-263. Frese, B., Moya, M., & Megias, J.L. (2004). So cial perception of rape: How rape myth acceptance modulates the influenc e of situational factors. Journal of Interpersonal Violence 19, 143-161. Fulero, S.M. & Delara, C. (1976). Rape victims and attributed responsib ility: A defensive attribution approach. Victimology, 1 551-563. Gabora, N.J., Spanos, N.P., & Joab, A. (1993). The effects of complainant age and expert psychological testimony in a simula ted child sexual abuse trial. Law and Human Behavior, 17 103-119. 237

PAGE 238

Gibbs, J.J. (1983). Problems and priorities: Percep tions of jail custodians and social service providers. Journal of Criminal Justice, 11, 327-338. George, W.H., & Martinez, L.J. (2002). Victim blaming in rape: Effects of victim and perpetrator race, type of ra pe, and participant racism. Psychology of Women Quarterly, 26, 110-119. Goodman, G.S., Bottoms, B.L., Herscovici, B.B., & Shaver, P. (1989). Determinants of the child victims perceived credibility. In S.J. Ceci, D.F. Ross, & M.P. Toglia (Eds.), Perspectives on childrens testimony. New York: Springer-Verlag. Goss, J.R., Peterson, K., Smith, L.W., Kalb, K., & Brodey, B.B. (2002). Characteristics of suicide attempts in a large urban jail system with an established suicide prevention program. Psychiatric Services 53, 574-579. Griffin, M.L. (2002). The influen ce of professional orientation on detention officers attitudes towards the use of force. Criminal Justice and Behavior 29, 250-277. Groth, A.N. & Burgess, A.W. (1980). Male rape: Offenders and victims. American Journal of Psychiatry, 137 806-810. Groves, R.M., Cialdini, R.B., & Couper, M.P. (1992). Understandi ng the decision to participate in a survey The Public Opinion Quarterly 56, 475-795. Guy, E., Platt, J.J., Zwerling, I., & Bullock, S. (1985) Mental health status of prisoners in an urban jail. Criminal Justice and Behavior 12, 29-53. Hartley, R.E. (1959). Sex-role pressures a nd the socialization of the male child. Psychological Reports, 5, 457-468. Hassine, V. (2004). Life without parole: Living in prison today (3rd ed.). Los Angeles: Roxbury. Hemmens, C. & Stohr, M.K. (2001). Correctional st aff attitudes regarding the use of force in corrections. Corrections Management Quarterly, 5, 27-40. Hensley, C., Koscheski, M., & Tewksbury, R. (2 005). Examining the characteristics of male sexual assault targets in a sout hern maximum-security prison. Journal of Interpersonal Violence 20, 667-679. Herek, G.M. (1984). Attitudes toward lesbians and gay men: A factor-analytic study. Journal of Homosexuality 10, 39-51. Herek, G.M. (1986). On heterosexual masculinity: Some psychical consequences of the social construction of gender and sexuality. American Behavioral Scientist, 29 563-577. 238

PAGE 239

Herek, G.M. (1988). Heterosexuals attitudes toward lesbians and gay men: Correlates and gender differences. The Journal of Sex Research, 25, 451-477. Herek, G.M. (2002). Heterosexuals attitudes toward bisexual men and women in the United States. The Journal of Sex Research 39, 264-274. Hodge, S., & Canter, D. (1998). Victims and perpetrators of male sexual assault. Journal of Interpersonal Violence 13, 222-239. Howard, J.A. (1984a). The normal victim: The effects of gender stereo types on reactions to victims. Social Psychology Quarterly 47, 270.-281. Howard, J.A. (1984b). Societal influences on attr ibution: Blaming some victims more than others. Journal of Personality and Social Psychology, 47, 494-505. Jackson, E., & Ammen, S. (1996). Race and correc tional officers punitive attitudes toward treatment programs for inmates. Journal of Criminal Justice 24, 153-166. Jackson, H. & Nuttall, R. (1994). Effects of gender, age, and history of abuse on social workers judgments of sexual abuse allegations. Social Work Research, 18 105-113. Jacobs, J.B. & Kraft, L.J. (1978). Integrating the keepers: A comparison of black and white prison guards in Illinois. Social Problems, 25, 304-318. Janoff-Bulman, R. (1979). Characterological vers us behavioral self-blame: Inquiries into depression and rape. Journal of Personality and Social Psychology 37, 1798-1809. Janoff-Bulman, R. (1982). Esteem and control base s of blame: Adaptive st rategies for victims versus observers. Journal of Personality 50, 180-191. Janoff-Bulman, R., Timko, C., & Carli, L.L. (198 5). Cognitive biases in blaming the victim. Journal of Experiment al Social Psychology, 21, 161-177. Jenkins, M.J., & Dambrot, F.H. (1987). The attribu tion of date rape: Obse rvers attitudes and sexual experiences and the dating situation. Journal of Applied Social Psychology, 17, 875-895. Jimenez, J.A. & Abreu, J.M. (2003). Race and sex effects on attitudinal perceptions of acquaintance rape. Journal of Counseling Psychology, 50 252-256. Johnson, R. (2002). Hard time: Understanding and reforming the prison (3rd ed.). Belmont, CA: Wadsworth Thompson. Jones, C., & Aronson, E. (1973). Attribution of fault to a rape victim as a function of respectability of the victim. Journal of Personality and Social Psychology 26, 415-419. 239

PAGE 240

Jones, E. F. & Forrest, J. D. (1992). Under-reporting of abortion in surveys of U.S. women: 1986-1988. Demography 29, 113-126. Jones, R. & Schmid, T. (1989). Inmates conceptions of prison sexual assault. The Prison Journal, 69, 53-61. Jordan, J. (2004). Beyond belief: Poli ce, rape, and womens credibility. Criminal Justice, 4, 2959. Junger-Tas, J., & Marshall, I.H. (1999). The se lf-report methodology in crime research. In M. Tonry (Ed.), Crime and Justice: A Review of Research 25, 291-367. Jurik, N.C. (1985). Individual and organizational determinants of correctional officer attitudes toward inmates. Criminology 23, 523-540. Jurik, N.C. & Halemba, G.J. (1984). Gender, working conditions and the job satisfaction of women in a nontraditional occupation: Female correctional officers in mens prisons. Sociological Quarterly, 25 551-566. Kanekar, S. & Kolsawalla, M.B. (1980). Responsib ility of a rape victim in relation to her respectability, attractiven ess, and provocativeness. Journal of Social Psychology, 112 153-154. Kassebaum, G.G., Ward, D.A., & Wilner, D.M. (1964) Some correlates of staff ideology in the prison. Journal of Research in Crime and Delinquency 1, 96-109. Kassing, L.R., & Prieto, L.R. (2003). The rape my th and blame-based beliefs of counselors-intraining toward male victims of rape. Journal of Counseling & Development, 81, 455461. Kauffman, K. (1981). Prison officers attitudes and perceptio ns of attitudes: A case of pluralistic ignorance. Journal of Research in Crime and Delinquency 18, 272-294. Kellar, M., Jaris, M., & Manboah-Roxin, J. (2001). T exas jail survey 2001: A status repor t. Austin: Texas Commission on Jail Standa rds. Retrieved September 15, 2008 from http://www.tcjs.state.tx.us/index.php?linkID=300 Kemper, E.A., Stringfield, S., & Teddlie, C. (1 998). Mixed methods samp ling strategies in social science research. In A. Tashakkori & C. Teddlie (Eds.), Handbook of mixed methods in social and behavioral research Thousand Oaks, CA: Sage. Kerle, K. (1982). The rural jail: Its people, problems and solutions. In S.D. Cronk J. Jankovic, & R.K.Green (Eds.), Criminal justice in rural America (pp. 189-204). United States Department of Justice, National Inst itute of Justice, Offi ce of Development, Testing, and Dissemination. 240

PAGE 241

Kerle, K. (1998). American jails: Looking to the future Boston: Butterworth-Heinemann. Kifer, M., Hemmens, C., & Stohr, M.K. (2003). The goals of corrections: Perspectives from the line. Criminal Justice Review, 28 47-69. Kimme, D.A. (1988). The nature of new small jails. American Jails, 1, 40-51. King, M., & Woollett, E. (1997). Sexually assa ulted males: 115 men consulting a counseling service. Archives of Sexual Behavior, 26, 579-588. Kinkade, P., Leone, M., & Semond, S. (1995) The consequences of jail crowding. Crime and Delinquency, 41, 150-161. Kleinke, C.L. & Meyer, C. (1990). Evaluation of rape victim by men and women with high and low belief in a just world. Psychology of Women Quarterly, 14 343-353. Klofas, J. (1986). Discretion among correctional offi cers: The influence of urbanization, age and race. International Journal of Offender Therapy and Comparative Criminology, 30, 111124. Klofas, J. (1990). The jail and the community. Justice Quarterly, 7, 69-102. Klofas, J., & Toch, H. (1982). The guard subculture myth. Journal of Research in Crime and Delinquency 19, 238-254. Kopper, B.A. (1996). Gender, gender identity, ra pe myth acceptance, and time of initial resistance on the perception of acquaintance rape blame and avoidability. Sex Roles, 34, 81-93. Koscheski, M., Hensley, C., Wright, J., & Tewks bury, R. (2002). Consensual sexual behavior. In C. Hensley (Ed.), Prison sex: Practice & policy (pp. 111-131). Boulder: Lynne Rienner. Krulewitz, J.E. (1981). Sex differenc es in evaluations of female and male victims responses to assault. Journal of Applied Social Psychology, 11 460-474. Krulewitz, J.E., & Nash, J.E. (1980). Effects of rape victim resistance, assa ult outcome, and sex of observer on attributions about rape. Journal of Personality, 47, 557-574. LaFree, G. (1981). Official reactions to social problems: Police deci sions in sexual assault cases. Social Problems, 28 582-594. LaFree, G.D., Reskin, B.F.,& Visher, C.A. (1985) Jurors response to vi ctims behavior and legal issues in sexual assault trials. Social Problems, 32 389-407. Lee, H.B. & Cheung, F.M. (1991). The attitudes to ward rape victims scale: Reliability and validity in a Chinese context. Sex Roles, 24 599-603. 241

PAGE 242

Lerner, M.J. (1970). The desire for justice and reactions to victims. In J. Macaulay & L. Berkowitz (Eds.), Altruism and helping behavior: Socia l psychological studies of some antecedents and consequences (pp. 205-229). New York: Academic. Lerner, M.J., & Miller, D.T. (1978). Just world research and the attri bution process: Looking back and ahead. Psychological Bulletin 85, 1030-1051. Levitt, E.E., & Klassen, A.K. (1974). Public atti tudes toward homosexuality: Part of the 1970 national survey by the Institute for Sex Research. Journal of Homosexuality, 1, 29-43. Little, R. (1983). The ignorable case, In W.G. Madow, I. Olkin, & D.B. Rubin (Eds.) Incomplete data in sample surveys (Vol.2). New York: Academic Press. Lockwood, D. (1980). Prison sexual violence New York: Elsevier. Maahs, J., & Pratt, T. (2001). Uncovering the pred ictors of correctional o fficers attitudes and behaviors: A meta-analysis. Corrections Management Quarterly, 5, 13-19. MacDonald, A.P., Huggins, J., Young, Y., & Sw anson, R.A. (1973). Attitudes toward Homosexuality: Preservation of sex morality or the double standard? Journal of Counseling and Clinical Psychology, 40, 161. MaCrae, C.N., & Shepherd, J.W. (1989). Sex differe nces in the perception of rape victims. Journal of Interpersonal Violence 4, 278-288. Mason, G.E., Riger, S., & Foley, L.A. (2004). The impact of past sexual experiences on attributions of res ponsibility for rape. Journal of Interpersonal Violence 19, 1157-1171. Masters, W. (1986). Sexual dysfunction as an aftermath of sexual assa ult of men by women. Journal of Sex & Marital Therapy 12 35-45. Mattick, H.W. (1974). The contem porary jails of the United Stat es: An unknown and neglected area of justice. In D. Glaser (Ed), Handbook of criminology (pp. 777-848). Chicago: Rand McNally. Maxfield, M.G. & Babbie, E. (2001). Research methods for crim inal justice and criminology (3rd ed.). Belmont, CA: Wadsworth. Mays, G.L. & Thompson, J.A. (1988). Mayberry revi sited: The characteris tics and operations of Americas small jails. Justice Quarterly, 5 421-440. McCaul, K.D., Veltum, L.G., Boyechko, V., & Crawfo rd, J.J. (1990). Understanding attributions of victim blame for rape: Sex, violence, and foreseeability. Journal of Applied Social Psychology, 20, 1-26. 242

PAGE 243

McFarland, S.T., Ellis, C. A., & Chunn, G. (2008). Report on rape in jails in the U.S .. Washington, DC: US Department of Justic e, Office of Justice Programs.Retrieved March 18, 2009 from http://www.ojp.usdoj.gov/reviewpa nel/pdfs/prea_finalreport_081229.pdf Mezey, G., and King, M. B. (1989). The effects of sexual assault on adult males: A study of 22 victims. Psychological Medicine 19, 205-210. Miller, E.E. (1978). Jail management: Problems, programs, and perspectives Lexington, MA: D.C.Heath and Company. Miller, R. (1982). A needs assessment for small ja ils in the United States In S.D. Cronk (Ed.), Criminal justice in rural America (pp. 205-229). Washington, DC: US Department of Justice, Office of Justice Programs, National Institute of Justice. Millham, J., San Miguel, C., & Kellogg, R. (1976) A factor-analytic c onceptualization of attitudes toward male and female homosexuals. Journal of Homosexuality, 2, 3-10. Mitchell, D., Hirschman, R., & Nagayama Ha ll, G.C. (1999). Attrib utions of victim responsibility, pleasure, and trauma in male rape. The Journal of Sex Research 36, 369373. Mori, L., Bernat, J.A., Glenn, P.A., Selle, L.L., & Zarate, M.G. (1995). A ttitudes toward rape: Gender and ethnic differences across As ian and Caucasian college students. Sex Roles, 32, 457-467. Murton, T.O. (1976). The dilemma of prison reform. New York: Holt, Rinehart and Winston. Nacci, P.L. & Kane, T.R. (1983). The incidence of sex & sexual aggressi on in federal prisons. Federal Probation, 7, 31-36. Nacci, P.L., & Kane, T.R. (1984). Sex and sexual aggression in federal prisons: Inmate involvement and employee impact. Federal Probation, 48 46-53. Nagel, B., Matsuo, H., McIntyre, K.P., & Morrison, N. (2005). Attitudes toward victims of rape. Journal of Interpe rsonal Violence, 20 725-737. National Institute of Corrections & The Moss Group, Inc. (2006). Staff perspectives: Sexual violence in adult prisons and jails Retrieved March 18, 2009 from http://nicic.gov/Library/021619 ODonnell, I. (2004). Prison rape in context. British Journal of Criminology, 44, 241-255. 243

PAGE 244

ODonohue, W., Elliott, A., Nickerson, M., & Valen tine, S. (1992). Perceived credibility of childrens sexual abuse alle gations: Effects of gende r and sexual attitudes Journal of Interpersonal Violence, 7, 147-155. ODonohue, W., Smith, V., & Schewe, P. (1998). The credibility of child sexual abuse allegations: Perpetrator gender a nd subject occupational status. Sexual Abuse, 10 17-24. Ohlander, J., Batalova, J., & Treas, J. (2005). Ex plaining educational influences on attitudes toward homosexual relations. Social Science Research 34, 781-799. OToole, M. (1997). Jails and prisons: The numbers say they are more diffe rent than generally Assumed. American Jails, 11, 27-31. Paternoster, R., Brame, R., Mazerolle, P, & Piquer o, A. (1998). Using the correct statistical test for the equality of regression coefficients. Criminology 6, 859-866. Perrott, S., & Webber, N. (1996). Attitudes towards male and female victims of sexual assault: Implications and services to the male victim. Journal of Psychology & Human Sexuality 8, 19-38. Petersilia, J. (2003). When prisoners come home: Parole and prisoner reentry New York: Oxford University Press. Pogrebin, M.R. & Poole, E.D. (1988). The work orientations of jail personnel: A comparison of deputy sheriffs and career line officers. Policy Studies Review, 7, 606-614. Pogrebin, M.R. & Poole, E.D. (1998). Women deputies and jail work. Journal of Contemporary Criminal Justice, 14, 117-134. Pollard, P. (1992). Judgements about victims a nd attackers in depicted rapes: A review. The British Psychological Society, 31, 307-326. Poole, E.D. & Pogrebin, M.R. (1988). Deputy sher iffs as jail guards: A study of correctional policy orientation and career phases. Criminal Justice and Behavior 15, 190-209. Poole, E.D. & Regoli, R.M. (1980). Role stress, custody orientation, and disciplinary actions: A study of prison guards. Criminology, 18, 215-226. Powell, T.A., Holt, J.C & Fondacaro, K.M. ( 1997). The prevalence of mental illness among inmates in a rural state. Law and Human Behavior, 21 427-438. Pugh, M.D. (1983). Contributory fault and rape c onvictions: Loglinear models for blaming the victim. Social Psychology Quarterly, 46, 233-242. 244

PAGE 245

Quackenbush, R.L. (1989). A comparison of androgynous, masculine sex-typed, and undifferentiated males on dimensions of attitudes toward rape. Journal of Research in Personality, 23 318-342. Rockefeller, J.D. (1972). Population and the American future. The report of the commission on population growth and the American future Retrieved July 15, 2007 from http://www.mnforsustain.org/rockefeller_1972_ glossary.htm Rubin, Z. & Peplau, L. (1975). W ho believes in a just world? Journal of Social Issues 31, 65-89. Ruddell, R. &Mays, G.L. (2007). Rural jails: Problematic inmates, overcrowded cells, and cash-strapped counties. Journal of Criminal Justice, 35, 251-260. Sabol, W.J. (2008a). Jail inmates at Midyear 2007 Bureau of Justice Statistics, US DOJ Office of Justice Programs, Retrie ved October 10, 2008 from http://www.ojp.usdoj.gov/bjs/pub/pdf/jim07.pdf Sabol, W.J. (2008b). Prison inmates at midyear 2007 United States Department of Justice, Office of Justice Programs, Bureau of Jus tice Statistics. Retrieved October 10, 2008 from http://www.ojp.usdoj.gov/bjs/pub/pdf/pim07.pdf Sabol, W.J., Minton, T.D., & Harrison, P.M. (2007) Prison and jail inmates at midyear 2006. United States Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. Retrieved July 15, 2007 from http://www.ojp.usdoj.gov/bjs/pub/pdf/pjim06.pdf Santos, M.G. (2004). About prison. Belmont, CA: Wadsworth Thomspon. Saum, C., Surratt, H., Inciardi, J., & Bennett, R. (1995). Sex in prison: Exploring the myths & realities. The Prison Journal, 75, 413-430. Sawyer, J. (1970). On male liberation. Liberation, 15, 32-33. Schmalleger, F. & Smykla, J. (2007). Corrections in the 21st Century (3rd ed.).Boston: McGraw Hill. Schneider, L.J., Soh-Chiew Ee, J., & Aronson, H. (1994). Effects of vict im gender and physical vs. psychological trauma/ injury on observer s perceptions of sexual assault and its aftereffects. Sex Roles 30, 793-808. Schuller, R.A., & Hastings, P.A. (2002). Compla inant sexual history evidence: Its impact on mock jurors decisions. Psychology of Women Quarterly, 26, 252-261. Schuller, R.A. & Stewart, A. (2000). Police responses to sexual assault complaints: The role of perpetrator/ complainant intoxication. Law and Human Behavior, 24 535-551. 245

PAGE 246

Schwarz, N., Strack, F., Hippler, H.J., & Bishop, G. (1991). The im pact of administration mode on response effects in survey measurement. Applied Cognitive Psychology 5, 193-212. S. Res. 1435, 108 Cong., (2003) (enacted). Shamir, B., & Drory, A. (1981). Some co rrelates of prison guards beliefs. Criminal Justice and Behavior, 8, 233-249. Sheldon, J.P. & Parent, S.L. (2002). Clergys attitudes and attributions of blame toward female rape victims. Violence against Women, 8 233-256. Smith, K.T. (1971). Homophobia: A tentative personality profile. Psychological Reports, 29, 1091-1094. Smith, N. & Batiuk, E. (1989). Sexual victim ization and inmate social interaction. The Prison Journal 69, 29-38. Smith, R. E., Keating, J. P., Hester, R. K., & Mitchell, H. E. (1976). Role and justice considerations in the attribution of responsibility to a rape victim. Journal of Research in Personality, 10, 346-357. Smith, R.E., Pine, C.J., & Hawley, M.E. (1988). So cial cognitions about adult male victims of female sexual assault. The Journal of Sex Research, 24, 101-112. Spears, J.W., & Spohn, C.C. (1996). The genuine vi ctim and prosecutors ch arging decisions in sexual assault cases. American Journal of Criminal Justice, 20 183-205. Stermac, L., del Bove, G., & Addison, M. (2004) Stranger and acquaintan ce sexual assault of adult males. Journal of Interpersonal Violence, 19, 901-915. Stephan, J.J., (2001). Census of jails, 1999 Washington, DC: US Departme nt of Justice, Office of Justice Programs, Bureau of Justice St atistics. Retrieved September 19, 2008 from http://www.ojp.usdoj.gov/bjs/pub/pdf/cj99.pdf Stephan, J.J. (2008). Census of state and federal correctional facilities, 2005 Washington, DC: US Department of Justice, Office of Justic e Programs, Bureau of Justice Statistics. Retrieved September 19, 2008 from http://www.ojp.usdoj.gov/ bjs/pub/pdf/csfcf05.pdf Stuart, I.R., & Greer, J.G. (1984). Victims of sexual aggression: Treatment of children, women, and men. New York: Van Nostrand Reinhold. Struckman-Johnson, C. & Struckman-Johnson, D. (1992). Acceptance of male rape myths among college men and women. Sex Roles 27, 85-100. Struckman-Johnson, C. & Struckman-Johnson, D. (2000). Sexual coercion rates in seven Midwestern prison facilities for men. The Prison Journal, 80, 379-390. 246

PAGE 247

Struckman-Johnson, C., & Struckman-Johnson, D. (2002). Sexual coerci on reported by women in three Midwestern prisons. The Journal of Sex Research, 39 217-227. Struckman-Johnson, C., Struckman-Johnson, D., Rucker, L., Bumby, K., & Donaldson, S. (1996). Sexual coercion reporte d by men and women in prison. The Journal of Sex Research, 33, 67-76. Studenmund, A.H. (2001). Using econometrics: A practical guide (4th ed.). Boston: Addison Wesley Longman. Sykes, G. (1958). The society of captives Princeton, N.J.: Princet on University Press. Tartaro, C. (2003). Suicide and the jail environment: An evaluation of three types of institutions. Environment and Behavior, 35, 605-620. Teplin, L.A. (1990). The prevalence of severe me ntal disorders among male urban jail detainees: Comparisons with the Epidemiologic Catchment Area program. American Journal of Public Health 80, 663-669. Teplin, L.A. (1994). Psychiatric and substance abuse disorders am ong male urban jail detainees. American Journal of Public Health 84 290-293. Teske, R.H.C. & Williamson, H.E (1979). Correcti onal officers attitude s toward selected treatment programs. Criminal Justice and Behavior, 6, 59-66. Tewksbury, R. (1989). Fear of sexual assault in prison inmates. The Prison Journal, 69, 6271. Tewksbury, R., & West, A. (2000). Research on se x in prison during the late 1980s and early 1990s. The Prison Journal, 80, 368-378. Thompson, E.H. & Pleck, J.H. (1986). The structure of male role norms. American Behavioral Scientist, 29 531-543. Thornberry, T.P., & Krohn, M.D. (2000). The self -report method for measuring delinquency and crime. In D. Duffee, D. McDowall, B. Ostrom R.D. Crutchfield, S.D. Mastrofski, & L.G. Mazerolle (Eds.), Measurement and analysis of crime and justice (pp. 33-83) Washington, DC: US Department of Justice. Thornton, B., Robbins, M.A., & Johnson, J.A. (1981). Social perception of the rape victims culpability: The influence of respondents personal-environmental causal attribution tendencies. Human Relations, 34, 225-237. Thornton, B. & Ryckman, R.M. (1986). The influence of a rape victims physical attractiveness on observers attributions of responsibility. Human Relations, 36, 549-562. 247

PAGE 248

Thornton, B., Ryckman, R.M., &Robbins, M.A. (1982). The relationships of observer characteristics to beliefs in the causal re sponsibility of victims of sexual assault. Human Relations, 35 321-330. Tjaden, P., & Thoennes, N. (2006). Extent, nature, and consequences of rape victimization: Findings from the national violence against women survey Washington, DC: US Department of Justice, Office of Justice Programs, National Institute of Justice. Retrieved September 10, 2007 from http://www.ncjrs.gov/pdffiles1/nij/210346.pdf Toch, H., & Klofas, J. (1982). Alienation and desire for job enrich ment among correction officers. Federal Probation 46, 35-44. Torrey, M. (1991). When will we be believed? Rape myths and the idea of a fair trial in rape prosecutions. University of California, Davis, Law Review, 24 1013-1071. Travis, J., Soloman, A.L., & Waul, M. (2001). From prison to home: The dimensions and consequences of prisoner reentry Washington, DC: The Urban Institute. Tucker, D. (1992). A punks song: View from the in side. In W. Dynes & S. Donaldson. (Eds.), Homosexuality & government, politics, & prisons (pp. 262-283). New York: Garland Publishing. Turner, C.F., Lessler, & Devore, J. (1992). Effects of mode of administration and wording on reporting of drug use. In C.F. Turner, J.T. Lessler, & G.C. Gfroerer, (Eds.), Survey measurement of drug use: Methodological studies (pp. 177-220). Washington, DC: Department of Health and Human Services. Van Voorhis, P., Cullen, F.T., Link, B.G., & Wolf e, N.T. (1991). The impact of race and gender on correctional officers orientation to the integrated environment. Journal of Research in Crime and Delinquency, 28, 472-500. Van Wormer, K. (1984). Becoming homosexual in prison: A socialization process. Criminal Justice Review, 9, 22-27. United States Census Bureau (1995). Urban and rural definitions Geography Division. Retrieved July 15, 2007 from http://www.census.gov/population/censusdata/urdef.txt United States Census Bureau (2000). Census 2000 geography glossary Retrieved July 15, 2007 from http://www.census.gov/ge o/www/tiger/glossry2.pdf United States Census Bureau (2008). County population estimates Retrieved March 4, 2009 from http://www.census.gov/popest/counties/ 248

PAGE 249

United States Department of Agriculture (2004). Measuring rurality: Rural-urban continuum codes. Economic research service. Retrieved July 15, 2007 from http://www.ers.usda.gov/brief ing/rurality/RuralUrbCon/ Wakelin, A., & Long, K.M. (2003). Effects of vic tim gender and sexuality on attributions of blame to rape victims. Sex Roles, 49, 477-487. Walker, J., Archer, J., & Davies M. (2005). Effects of rape on men: A descriptive analysis. Archives of Sexual Behavior, 34, 69-80. Weis, K., & Borges, S.S. (1975). Victimology and rape : The case of the legitimate victim. In L. Schultz (Ed.), Rape victimology (pp. 91-141). Springfield, IL: Charles C. Thomas. Weisburd, D., Petrosino, A., & Mason, G. (1993). Design sensitivity in criminal justice experiments. In M. Tonry (Ed.), Crime and Justice: A Review of Research, 17 337-379. Weiss, C. & Friar, D. (1974). Terror in the prisons: Homosexual rape & why society condones it. Indianapolis: Bobbs-Merrill. Welsh, W.N. (1992). The dynamics of jail reform litig ation: A comparative an alysis of litigation in California counties. Law and Society Review, 26 591-625. West, H.C. & Sabol, W.J. (2008). Prisoners in 2007. United States Department of Justice, Office of Justice Programs, Bureau of Jus tice Statistics. Retrie ved March 18, 2009 from http://www.ojp.usdoj.gov/bjs/pub/pdf/p07.pdf Whatley, M.A. (2005). The effect of participant sex, victim dress, and traditional attitudes on causal judgments for marital rape victims. Journal of Family Violence, 20 191-200. Whatley, M.A., & Riggio, R.E. (1993). Gender differences in attributions of blame for male rape victims. Journal of Interpersonal Violence, 8, 502-511. White, B.H., & Robinson Kurpius, S.E. (2002). E ffects of victim sex and sexual orientation on perceptions of rape. Sex Roles 46, 191-200. Whitehead, J., Linquist, C., & Klof as, J. (1987). Correctional officer professional or ientation: A replication of the Klofas-Toch measure. Criminal Justice and Behavior, 14, 468-486. Wooden, W. & Parker, J. (1982). Men behind bars: Sexual exploitation in prison New York: Plenum. Wortley, R. (2002). Situational prison contro l: Crime prevention in correctional institutions Cambridge, MA: Cambridge University. 249

PAGE 250

250 Xenos, S. & Smith, D. (2001). Perceptions of rape and sexual assault among Australian adolescents and young adults. Journal of Interpersonal Violence, 16 1103-1119.

PAGE 251

BIOGRAPHICAL SKETCH Carrie Lynn Cook received her Bachelor of Ar ts degree from the University of North Carolina at Greensboro in 1999. After majoring in sociology as an undergraduate, she enrolled in the Master of Science program at Valdosta St ate University. After completing the Master of Science in criminal justice in 2002, she worked as a correctional counselor for the Georgia Department of Corrections for three years. She enrolled in the Doctoral program at the University of Florida in Criminology, Law, and So ciety in 2005. Her rese arch interests include corrections, sentencing policy, victimization, a nd program evaluation. Carrie has accepted a position in the Department of Government and Sociology at Georgi a College and State University in Milledgeville, Georgia.