SEXUAL REPEAT VICTIMIZATION AND TRAUMA : AN ECOLOGICAL PERSPECTIVE by ELIZABETH ASHLEY TOMSICH B.A., University of Minnesota, 2003 M.S., Purdue University, 2007 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Affairs and Administration 2013
ii This thesis for the Doctor of Philosophy degree by Elizabeth Ashley Tomsich has been approved for the Public Affairs and Administration Program by Angela Gover, Chair Mary Guy Callie Rennison Bonnie Fischer Date : 6/11/13
iii Tomsich, Elizabeth Ashley (Ph.D., Public Affairs and Administration ) Sexual Repeat Victimization and Trauma : An Ecological Framework Thesis directed by Professor Angela R. Gover. ABSTRACT Sexual repeat victimization, or more than one sexual assault perpetrated by distinct offenders is a well documented social problem Repeat victimizations disproportionately comprise a large number of tota l incidents, and survivors of se xual victimization are 2 to 3 times more likely to be revictim ized than those never sexually assaulted (Merrill Newel l, Gold, & Milner 1999; Russell, 1986). Survivors of sexual repeat victimization are more likely to develop depression, anxiety, posttraumatic stress disorder (PTSD), and substance addictions than those sexually assaulted by one perpetrator only and the g eneral population D evelopment of an evidence-based policy response to sexual repeat victimization and post -victimization psychological trauma is critical However, the origins of sexual repeat victimization and its long term impact on mental health are un clear To date, research has not established a reliable theory of sexual repeat victimization that distinguishes between those who are victimized by one perpetrator and those repeat victimized by multiple perpetrators The absence of a n explanatory sexual repeat victimization theory may be due to studies stratified emphases on psychological, behavioral, and situational risk and protective factors. With respect to post -victimization mental health although sexual repeat victimization is a documented risk factor for depression, few studies have modeled mental health outcomes for those victimized by one perpetrator and repeat sexual victims separately The current research contributes to the sexual repeat victimization literature through employing a mod el of repeat rape
iv victimization that integrates individual, situational, and sociocultural variables in a national probability sample In addition, unique risk and protective factors for long -term depression among those sexually victimized by one perpetrat or and repeat rape victims are identified Research questions addressed are: 1) Which ecological factors predict respondents victimized by one perpetrator only versus repeat victims of sexual violence? 2) Which ecologica l factors predict depression among respondents victimized by one perpetrator only versus repeat victims of sexual violence? Data are drawn from Tjaden and Thoennes (1998 ) National Violence Against Women Survey (NVAWS). Binary logistic regression identifies predictors of sexual repeat victi mization and OLS regression distinguishes indicators of depression for respondents surviving rape by one perpetrator versus repeat victimizations by separate perpetrators F indings are discussed with consideration of victimization theory and policy implica tions The form and content of this abstract are approved. I recommend its publication. Approved: Angela R. Gover
v DEDICATION To my parents, Katherine and Peter Tomsich, who inspired me to pursue higher education and whose enduring support and encouragement made this all possible
vi ACKNOWLEDGMENTS My dissertation journey has successfully concluded thanks to the support and wisdom of a number of individuals Dr. Angela Gover s dedication to overseeing this dissertation and my development as a scholar was and is truly incomparable Dr. Govers decision to become my advisor and believe in me has fatefully shaped my pathway for the better and I am truly fortunate to have experienced the privilege of her mentorship In addition, I would like to th ank my dissertation committee members of Callie Rennison, Mary Guy, and Bonnie Fisher for their guidance collegiality, and thoughtful recommendations as I learned through multiple trials and errors how an idea culminates into a completed research project. I must also acknowledge the lessons imparted to me by my personal and academic support systems I thank Dr. Angela Gover for teaching me that mentorship is one of the most powerful, life changing gifts a person can give to another I thank my family for teaching me the power of unconditional support in prevailing over seemingly insurmountable obstacles and setbacks. I thank Jennifer Bruce for demonstrating that one woman can function as a comprehensive emergency support system, whether that manifests as a tire change, statistical assistance, or a beer pitcher companion. I thank Jennifer Kluckner for inexplicably loving me throughout my monogamous dedication to my computer over the last five years and for teaching me that love is a mighty weapon in sl aying academic dragons Lastly, I thank everyone for modeling endless patience and understanding in the face of multiple reported apocalypse s, each of which turned out merely to be storm s
vii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................................................................................................ 1 Contribution of Research ............................................................................................ 3 Scope of the Problem ................................................................................................... 4 Consequences ............................................................................................................... 8 Organization ................................................................................................................. 9 II. SEXUAL REPEAT VICTIMIZATION .......................................................................... 11 History of Victimology ............................................................................................. 11 Early History of Victimology ............................................................................. 11 Definitions of Victim ....................................................................................... 13 Victimology: Scientific Inquiry and Political Response ................................... 14 Victimology: Philosophies of Science ............................................................... 15 Models on the Ori gins of Victimization ............................................................ 17 Situation -Oriented Victimization Theories ................................................. 17 Social -Structural and Cultural Victimization Theories ............................... 19 Conclusion ........................................................................................................... 20 Theoretical Background: Models of Sexual Re peat Victimization ........................ 21 Psychological and Trauma Factors ..................................................................... 21 Behavioral Factors ............................................................................................... 27 Situational Factors ............................................................................................... 28 Summary .................................................................................................................... 30
viii III. POST SEXUAL VICTIMIZATION MENTAL HEALTH .......................................... 32 Theoretical Background: Models of Post -Sexual Victimization Mental Health ... 32 Victim Characteristics ......................................................................................... 33 Context of the Sexual Assault ............................................................................. 36 Post -Sexual Assault Experience ......................................................................... 39 Summary .............................................................................................................. 44 IV ECOLOGICAL FRAMEWORK .................................................................................... 45 Framework Assumptions........................................................................................... 47 Ecological Risk and Protective Factors for Sexual Repeat Victimization and Trauma ................................................................................................................. 49 Childhood Development ............................................................................... 49 Sexual Repeat Victimization .................................................................. 49 Post -Sexual Victimization Trauma ........................................................ 50 Context of Victimizat ion .............................................................................. 50 Sexual Repeat Victimization .................................................................. 50 Post -Sexual Victimization Trauma ........................................................ 51 Social Power .................................................................................................. 52 Sexual Repeat Victimization .................................................................. 52 Sexual Victimization Trauma ................................................................. 53 Implications of Ecological Models of Sexual Repeat Victimization and Post Sexual Victimization Trauma ............................................................................. 53 Summary .................................................................................................................... 54 V RESEARCH METHODS ................................................................................................. 56 Research Questions .................................................................................................... 56 Study Design .............................................................................................................. 56 Sample .................................................................................................................. 58
ix Measures .............................................................................................................. 63 Dependent Variables ..................................................................................... 63 Research Question One: Predictors of Rape Victimization by One Perpetrator Versus Multi ple Victimizations by Separate Perpetrators 63 Research Question Two: LongTerm Post -Victimization Mental Health ....................................................................................................... 65 Independent Variables .................................................................................. 66 Childhood Development ......................................................................... 67 Context of Rape Victimization Variables .............................................. 70 Social Power ............................................................................................ 71 Control Variables .................................................................................... 73 Data Analysis ............................................................................................................. 74 Missing Data Analysis ........................................................................................ 75 Selection Bias ...................................................................................................... 77 Summary .................................................................................................................... 80 VI. RESULTS ......................................................................................................................... 82 Sample Description .................................................................................................... 82 Demographic Variables ....................................................................................... 83 Rape Victimization Variables ............................................................................. 84 Research Question One ............................................................................................. 86 Research Question Two ............................................................................................. 88 Post -Victimization Mental Healt h among All Rape Victims: Bivariate Relationships ........................................................................................................ 88 Post -Victimization Mental Health among Respondents Sexually Victimized by One Perpetrator Only: Bivariate Relationships ............................................ 90 Post -Victimization Mental Health among Respondents Sexually Victimized Multiple Times by Separate Offenders: Bivariate Relationships ..................... 91
x Post -Victimization Mental Healt h among All Rape Victims: Multivariate Relationships ........................................................................................................ 93 Post -Victimization Mental Health for Respondents Sexually Victimized by O ne Perpetrator Only: Multivariate Relationships ............................................ 94 Post -Victimization Mental Health for Respondents Sexually Victimized Multiple T imes by Separate Offenders: Multivariate Relationships ................ 95 Summary .................................................................................................................... 97 VII. DISCUSSION ................................................................................................................. 99 Research Question One ............................................................................................. 99 Control Variables ................................................................................................. 99 Childhood Development ..................................................................................... 99 Context of Rape Victimization ......................................................................... 100 Social Power ...................................................................................................... 104 Research Question Two ........................................................................................... 106 Control Variables ............................................................................................... 106 Childhood Development ................................................................................... 106 Context of Rape Victimization Variables ........................................................ 107 Social Power ...................................................................................................... 109 Research Limitations ............................................................................................... 110 Future Research ....................................................................................................... 113 Alternative Validated Measures ....................................................................... 115 Sexual Repeat Victimization Model Comparison Research ........................... 116 Policy Implications .................................................................................................. 120 Conclusion................................................................................................................ 125
xi BIBLIOGRAPHY ................................................................................................................ 128 APPENDIX .......................................................................................................................... 150 A. TABLES ...................................................................................................... 150
xii LIST OF TABLES Table 5.1 : Sample Characteristics. .................................................................................................................................... 6 0 5.2: Ecological Variables in Sexual Repeat Victimization Likelihood Model....67 5.3: Variables Ecological in P ost -victimization Mental Health Model......................67 6.1: Descriptive Statistics for Sample and Bivariate Comparisons between One Perpetrator Victims and R epeat Rape V ictims (n=1381)..............8 1 A.1: Percentag e of Missing Data on Incomplete C ases....................................................... .1 49 A.2: Logistic Regression for Likelihood of Repeat Victimization Model..........1 51 A.3: Pre -Victimization Cross Tabulation T ables.......................................... ................. ...........1 52 A.4: OLS Regression of LongTerm Post -Victimization Mental Health. ...........1 53
xiii LIST OF FIGURES Figure 4.1 : Conceptual Groupings of Variables in the Ecological Frame work...........................4 7 5.1 : Oper ationalization of Sexual Repeat Victimization...................................................6 5
1 CHA PTER I INTRODUCTION T he phenomenon of repeat victimization, or multiple criminal victimizations experienced by one person has been documented both across and within crime types (Farrell & Pease, 2001; Finkelhor, Ormrod, & Turner, 2007). R epeat sexual assault receives particular academic focus (Fisher, Daigle, & Cullen, 2009; 2008; Daigle, Fisher & Cullen, 2008), and is found among male and female samples (Desai, Arias, Thompson, & Basile, 2002). In a recent study of college students, t wenty -three perc ent of rape victims reported more than one rape during one school year (Daigle et al., 2008) Women reporting three or more sexual victimizations of any kind, 3.3% of the sample, experienced almost half (45.2%) of all sexual assaults (Daigle et al., 2008). Another study indicate s that female university students sexually abused in childhood are more than twice as likely (63.6%) to experience a subsequent sexual assault or rape in college than those never victimized in childhood (26.6%) (Gidycz, Hanson, & Layman, 1995). Experiencing s exual repeat victimization escalates risk for depression, anxiety, Post Traumatic Stress Disorder (PTSD) addiction, and violent offending among survivors ( Daigle et al., 2008; Hosser, Raddatz, & Windzio, 2007; Kimerling Alvarez, Pavao, Kaminski, & Baumrind, 2006). Presently interpersonal violence policies targeting domestic and sexual violence often focus on particular incidents of violent victimization (Kilpatrick, 2004), preventing the initial incident, or the immedia te post incident response (Daigle et al., 2008). Many rape crisis centers and battered women's shelters gather intake data about individual victimizations, but may not collect comprehensive information on a broad -spectrum violence history (Kilpatrick, 2004). Likewise, college
2 sexual assault interventions commonly emphasize reporting and providing medical and psychological support for the most recent incident (Daigle et al., 2008). Such policies provide critical assistance to victims in the aftermath of victimization, but are unresponsive to the risk of future victimizations by different perpetrators, and overlook the role of cumulative trauma in psychological distress The threat of repeat victimization interfering with post -sexual assault recovery is disqui eting from both a normative and public policy perspective. Equally, the need to support sexual violence victim recovery programs that address lifetime trauma is apparent. However to date, research has not consistently verified the variables and processes predicting sexual repeat victimization and long term recovery, hindering the development of an evidence -based policy response (Macy, 2008). As a result the present research exa mines multiple indicators that may act as risk or protective factors for sexual repeat victimization and long term post -victimization distress. Although sexual repeat victimization theories examining victim characteristics and lifestyles are prevalent in l iterature, the current study takes an ecological approach to assess indi vidual, situational, and socio cultural factors influence on sexual repeat victimization and post -victimization mental health Multivariate analys es answer the following research quest ions: 1) Which ecologically based factors predict respondents victimized by one perpetrator only versus repeat victims of sexual violence? 2) Which ecologically based factors predict depression among respondents victimized by one perpetrator only versus repeat victims of sexual violence?
3 Contribution of Research The current research contributes to the literature on sexual repeat victimization and post -victimization trauma in several respects. Extant research on sexual repeat victimization risk and protective factors largely examines re peat victimization among college and clinical populations who first experienced sexual abuse in childhood (Arata, 2002) and separately assesses psychological, behavioral, and situational variables ( Fargo, 2009). Prior research has contributed prospective examinations of sexual repeat victimization risk and protective factors through continued access to special clinical and university student populations ( Fargo, 2009; Himel ein, 1995; Katz, May, Srensen, & DelTosta, 2010). The present research complements longitudinal research findings among special populations through examining rape victimization at any point in development in a large, national random digit -dialing cross -sectional sample. In addition, the current research employs a model of repeat sexual victimization that integrates ecologically based individual, behavioral, situational, and sociocultural variables in one study. In contrast to sexual repeat victimization likelihood research, several scholars have proposed and assessed ecological models of post -sexual assault recovery and distress (Charuvastra & Cloitre, 2008 Neville, Heppner, Oh, Spanierman, & Clark, 2004; Samuels Dennis, Ford Gilboe, Wilk, Avison, & Ray, 2010). Previous research has investigated short term post -victimization distress of survivors in clinical and college samples (Resnick, Walsh, Schumacher, Kilpatrick, & Acierno, 2013) and psychological functioning longer term in community an d national samples (Classen, Field, Koopman, Nevill -Manning, & Spiegel, 2 001; Najdowski & Ullman, 2011; Ullman & Brecklin
4 2002a, 2002b; Ullman, Filipas, Townsend, & Starzynski, 2006; Wyatt, Guthrie, & Notgrass; 1992; Zinzow, Resnick, Amstadter, McCauley, R uggiero, & Kilpatrick, 2010). However, the majority of research considers post assault recovery with respect to the most recent or traumatic sexual assault, with repeat victimization as a risk factor. The present research contributes to the body of literature on long -term post -sexual assault recovery through examining differentiated models of risk and protective factors for depression symptoms among respondents in a large national probability sample reporting rape victimization by one perpetrator versus repeat rape victimization by multiple offenders. Scope of the Problem Rates on the prevalence of experiencing sexual assault, or sexual touching or intercourse by any form of coercion, over ones lifetime vary from 7.3% (Norris, 1992) to 16.7% (Burnam et al., 1988) among college and general population samples. Estimates fluctuate depending on measures used and rates may almost double if attempted sexual assaults are included (Koss, Gidycz, & Wisniewski, 1987). In o ne landmark study on sexual assault incidents researchers conduct ing face to -face interviews f ound that 44% of the sample report ed a completed (24%) or attempted sexual assault (20%) (Russell, 1986 ). Sexual assault is primarily committed against the young by perpetrators known to the victims, and it remains a rarely reported crime (Tjaden & Thoennes, 1998). In addition, survivors share a common threat: a heightened likelihood of future, repeat sexual victimization by another perpetrator Research on rates of repeat victimization across crime types finds the scope to be considerable. Around 13.5% of respondents in a national Canadian survey reported more
5 than one violent or property crime victimization, and repeat victimizations comprised 54% of all incidents (Gabor & Mata, 2004). Early British Crime Survey data analysis found that 14% of respondents reported 70% of the victimizations (Gottfredson, 1984). Assessment s of repeat violent crime victimization document up to two -thirds of victims e xperience more than one sexual or physical assault throughout their lifetime (Kilpatrick, Resnick, Saunders, & Best, 1998). Women in Kilpatrick et al. s (1998) sample who reported previous physical or sexua l assault were five times more likely to experienc e a new assault, even after controlling for age race, education, and substance use risk factors Over the two years following initial reporting, 10.8% of the women with a history of two sexual or physical assaults experienced a new victimization, and 23% of women who had experienced three or more victimizations experienced a new sexual or physical assault (Kilpatrick et al., 1998) College students were especially vulnerable to new victimizations compared with non -students (10.2% vs. 4.0%) (Kilpatrick et a l., 1998). Likewise, women in violent intimate relationships experience high rates of ongoing polyvictimization with o ne -third to one -half of all women in physically violent relationships also sexually assaulted by their partners ( Frieze & Browne, 1989). With consideration of relative risk, w omen sexually abused in childhood are two to five times more likely to experience later sexual or physical victimization commonly before high school ends, compared with women who do not experience childhood sexual abuse (Coid Petruckevitch, Feder, Chung, Richardson, & Moorey, 2001; Messman-Moore & Long, 2000; Noll, Horowitz, Bonanno, Trickett, & Putnam, 2003; West, Williams, & Siegel, 2000; White & Widom, 2003). The risk of repeat physical or sexual assault is even greater among samples repor ting childhood sexual abuse and/ or childhood maltreatment.
6 Kimerling et al. (2007) found that those victimized in childhood by physical or sexual abuse had a 5.8 times greater likelihood of adult physical or sexual victimization than those never vi ctimized in childhood. In terms of sexual repeat victimization specifically, Burnam et al. (1988) documented in a community sample that two thirds of respondents with a history of sexual assault, or pressured or forced sexual contact, r eported two sexual assaults throughout their lifetime. The majority of first-time assaults (80%) were perpetrated against the young, when respondents were between 6 and 25 years old (Burnam et al., 1988). Relative risk is substantial, with s urvivors of chi ldhood sexual assault 2 to 3 times more likely to be sexually assaulted or raped in adulthood than women who do not experience childhood sexual abuse (Maker et al. 2001; Merrill et al., 1999; Russell, 1986; West et al. 2000). A meta analysis of sexual repeat victimization literature documents an effect size of .59 between childhood sexual abuse and sexual repeat assault or rape in adulthood (Roodman & Clum, 2001). Variations in reports of the respective risk of future sexual victimization for non -victim s and victims often relate to differences in sampling and definitions of sexual victimizati on (Roodman & Clum, 2001). Consequently, the subsequent review of repeat victimization risk research is discussed with consideration of sampling and operational izati on of sexual victimization Van Bruggen, Runtz, and Kadlec (2006) examined a Canadian university sample for experiences with childhood sexual assault, or sexual contact before the age of 14 by force or by someone five or more years older Women with a hist ory of childhood sexual assault had twice the odds of experiencing sexual assault, attempted rape, or rape after the age of 14 in contrast to those who had not been sexually victimized in childhood (Van
7 Bruggen et al., 2006) Another study examining childhood or adolescent sexual assault and repeat adult sexual assault in a college sample found that 29.4% of survivors reported a later sexual victimization during a one semester period versus only 11.9% of the women who were not sexually abused in childhood (Reese-Weber & Smith, 2011). Research conducted in community samples finds similar rates of sexual repeat victimization. Wyatt et al. (1992) studied a Los Angeles County random -digit dialing sample regarding women's experiences with nonconsensual exposure to masturbation or someones genitals or attempted or completed rape I ncidents occurring prior to the age of 18 were categorized as childhood sexual abuse, and those after the age of 18 were considered adult sexual assault W omen reporting sexu al abuse in childhood were 2.4 times more likely to be sexually assaulted in adulthood (Wyatt et al., 1992) Among women who experienced contact sexual abuse, 55.7% were sexual assaulted in adulthood versus only 21% of those who did not experience contact childhood sexual abuse (Wyatt et al., 1992). In another community sample, Najdows ki and Ullman (2011) recruited Chicago residents through newspaper announcements and flyers, inviting women ages 18 and older with unwanted sexual experiences to participate in a paid research study over a one year period. The researchers assessed sexual assault, or unwanted sexual contact, sexual coercion, or attempted or completed rape, since the age of 14 or older, at Time 1 and one year later at Time 2. Forty -five percent of the women reported a sexual assault at the second wave of measurement (Najdowski & Ullman, 2011). Casey and Nurius (2005) used random -digit -dialing to collect information on forced sexual contact and attempted or completed rape among women in Washington State (Casey & Nurius, 2005). Thirty -
8 eight percent of the sample reported at least one experience of sexual assault over their lifetime (Casey & Nurius, 2005). Of that 38%, almost half experienced a single incident, one third experienced multiple victimizations by the same perpetrator over time, and one fifth experienced repeat victimization, or multiple sexual assaults by different perpetrators over time (Casey & Nurius, 2005). Research conducted in both college and community samples indicate a si gnificant proportion of sexual assault victims experience repeat sexual victimization, and that victimization is strongly associated with the likelihood of future victimizations. The incidence of sexual repeat victimization not only provides researchers wi th an estimate of population repeat victimization rates but it also informs mental health providers on the extent of cumulative post -victimization trauma in communities Just as any sexual victimization is associated with greater mental health distress th an the general population; repeat sexual victimization predict s more severe mental health issues than victimization by one perpetrator alone. Consequences Individuals sexually victimized multiple times face profound psychological trauma Respondents in Banyard, Williams, and Siegel s (2001) sample with a history sexual or physical assault in both childhood and adulthood indicate d greater problems on a range of mental health subscales, including anxiety, depression, dissociation, sexual concerns, dysfunct ional sexual behavior, intrusions, defensive avoidance, and impaired self reference, than those victimized only in chil dhood or never victimized Kimerling et al. (2007) observed that repeat victims of physical or sexual assault had a higher risk of anxiet y, depression, and PTSD than respondents who were victimized only in childhood
9 Other research in a community recruited sample documented that women who survive d repeat sexual contact sexual coercion attempted rape, or completed rape describe d greater in terpersonal problems than those not revictimized and non -victims (Classen et al., 2001) Likewise, c ollege women reporting contact childhood sexual abuse and later attempted or completed rape presented more depression, anxiety, hostility, and PTSD symptoms than survivors of childhood sexual abuse only in Messman -Moore, Long, and Siegfrieds (2000) research. However, higher rates of trauma among repeat victims may originate in factors other than victimization history alone. S ome research suggests pos t -sexual victimization recovery accounts for a portion of the variance in the relationship between victimization by one perpetrator and repeat victimization. Arata (2000) assessed post -childhood sexual abuse functioning through measuring PTSD symptoms, num ber of consensual sexual partners, and the degree to which respondents felt responsible for the childhood sexual abuse in a sample of female undergraduates. Victim self blame, posttraumatic symptoms, and sexual activity were found to mediate the associatio n between childhood sexual abuse and repeat victimization. Mental health problems that emerge as a result of sexual victimization are compounded by the experience of repeat victimization. Consequently, a considerable portion of victims are not only at risk of being targeted by distinct perpetrators for repeat sexual victimization, but are additionally at risk to suffer serious psychological problems and disorders as a result of cumulative trauma Organization Chapter 1 introduced the current researchs contribution to the literature and the scope and consequences of sexual repeat victimization Expanding upon the literature
10 presented in Chapter 1, Chapter 2 reviews models of sexual repeat victimization in three sections. The first sec tion discusses psychological and trauma models of sexual repeat victimization. The second and third sections review the more finite literature on behavioral and situational models of sexual repeat victimization. Chapter 3 addresses post -sexual victimizatio n distress theories and risk and protective factors with respect to three conceptual categories: 1) victim characteristics, 2) context of the assault, and 3) post assault experience. Chapter 4 introduces the framework of the current research, the ecologica l model. After reviewing the components and assumptions of the ecological model, this framework is applied to sexual repeat victimization risk and post -sexual victimization distress among respondents victimized by one perpetrator only versus repeat victims of rape by separate perpetrators With the literature reviewed and theoretical groundwork positioned, Chapter 5 introduces the current studys research questions, sample, measures, and analytic approach. Results of the analytic examinations of both resear ch questions are then considered The present research concludes in Chapter 6, with a discussion of the analytic findings with respect to previous research, methodological limitations, future research, and policy implications
11 CHAPTER II SEXUAL REPEAT VICTIMIZATION History of Victimology In order to contextualize repeat victimization research within the broader victimization literature, a brief history of victimology will be reviewed. Victimology emerged in the 1940s as a sub-discipline of criminology, where research on victimization, most broadly defined as injury or adversity by any man-made cause, was argued to be an important empirical complement to the study of criminality. Although the roots of victimology were arguably tainted by prejudicial noti ons of victim guilt and true innocence, scholarship has progressed to include macro -level theories alongside enduring micro level and interactionist models of victimization. Early History of Victimology Whereas criminologists investigate why certain peopl e are motivated to engage in criminal behavior, victimologists research why particular individuals are identified by offenders as ideal targets to victimize (Karmen, 2 012). Until the 1940s and 1950s criminology focused on offenders, the justice system's re sponse to offenders, and the balance between societal safety and offender rehabilitation (Karmen, 2012). Although the term victimology is inelegant (Sparkes, 1982, pp. 22) at best and evocative of victim stigma at worst, this field may be alternately tit led survivorology (Karmen, 2012). V ictimologists are viewed as researching vulnerability to victimization However, victimology may be reinterpreted as the study of resiliency in survivors of crime, particularly among those who experience multiple or ongoing victimizations (Karmen, 2012).
12 The genesis of victimology as a field is commonly accredited to Mendelsohn (1940), von Hentig (1940, 1948), Nagel (1963), Ellenberger (1955), and Wertham (1949). Other scholars have referred to works by Beccaria in the 18th century and Garafalo (1914) as earlier forefathers of victimology (OConnell, 2008; Kirchhoff, 2003). Kirchhoff (2003) calls attention to Beccarias writings on the notion of u nreasonable laws, where power is granted to a privileged few and divested from the majority. Another early contributor is Garafalo (1914), who discussed the concept of victim provocation in his works (OConnell, 2008). The first reference to the word "vic timology" has been alternately attributed to Benjamin Mendelsohn in 1947, or to Werthems (1949) book The Show of Violence proposing scientific inquiry on victims (Dussich, 2003; Fattah, 2000; Kirchhoff; 2006; OConnell, 2008). Works by Mendelsohn (1940) a nd Von Hentig (1940) are considered to be emblematic of early victimology. Mendelsohn, a defense lawyer, interviewed rape victims to develop a 6-category victim typology with only one, the "innocent," considered truly devoid of responsibility for her vict imization Mendelsohn later established General Victimology which assessed not only assault, but also car accidents, disasters, and genocide (Dussich, 2003). In addition, Mendelsohn advocated for victims rights, criticizing the criminal justice system s t reatment of victims, and recommending the development of victim assistance centers and research institutes (Karmen, 2012). Von Hentig (1940) similarly constructed a typology of victims based on victim proneness, of which three were most common: depressive, acquisitive, and wanton. The depressive type is nave and careless, the acquisitive victim is manipulated due to their own motivations weakening threat perception, and the wanton type is vulnerable to
13 developmental stage stressors. Mendelsohn (1940) and von Hentig (1940) suggested that many victims deliberately or unintentionally influence their victimization through provoking the criminal or placing themselves in unsafe situations, engaging as a penal couple (Mendelsohn, 1963) or in a duet with the of fender (von Hentig, 1948). However, both scholars acknowledge d that numerous crimes are committed where the victim did not reciprocally contribute to the perpetration of the crime (Fattah, 2000). Von Hentig (1948) later published The Criminal and His Victi m challenging criminology's circumscribed focus on the criminal (Fattah, 2000). Several scholars extended Mendelsohn (1940) and von Hentigs (1940) perspectives through studying homicide victim precipitation (Wolfgang, 1957), victims experiences with the criminal justice system (Nagel, 1963), and rape victim precipitation (Amirs, 1967). Backlash to early victimology began in the late 1960s and 1970s, particularly regarding rape victim precipitation (OConnell, 2008). M any scholars additionally critiqued early victimology for lacking legitimacy as an empirical and theoretical field (Rock, 2002). Critics stressed the 20th century gender norms evident in victim precipitation approaches, and the lack of the scientific method in both Mendelsohn (1940) and Von H entigs (1940) works (OConnell, 2008). In spite of these legitimate concerns early victimologists writings laid the groundwork for contemporary victimology, instigating a paradigm shift in criminology to consider both the offender and the victim (OConn ell, 2008). Definitions of Victim The World Society of Victimology and Karmen (2012) consider victimology to be the scientific examination of crime victims, scope and origins of victimization,
14 aftermath for those involved, and societal and individual responses to victims of crime. Although Fattah (1991) acknowledges sources of victimization beyond crime, such as natural disasters, he suggests victimology should focus on victims of crime (OConne ll, 2008). Alternately, Kirchhoff (2003) exercises a broader definition similar to Mendelsohn (1963), including any man-made victims or victims of human rights violations. Knudten (1992) suggests crime, disasters, accidents, and abuse of power victimizatio ns are separate but overlapping categories (OConnell, 2008). Victimology: Scientific Inquiry and Political Response Although key figures in the early 20th century contributed to criminologys expanded focus on both offenders and victims, scientific inquiry on victimization did not emerge until after the Second World War (Fattah, 2000). During the 1960s, the post -war baby boom contributed to an increase in street crime, increasing pressure on the government to serve the needs and rights of victims (Karmen, 2012). In addition, the feminist movement of the 1960s and 1970s raised awareness on the epidemic of interpersonal violence, and the degree to which society failed to protect women and children The enactment of California's Victi m Compensation program i n 1965 set the precedent for state -by-state diffusion of victim restitution policies (Dussich, 2003). Subsequently in 1985, the United Nations adopted the Declaration of Basic Principles of Justice for Victims, and many American states followed suit (Fatta h, 2000). State statutes commonly mandated the right s of victims to be notified and participate in judicial prosecution, to receive recovered stolen property, to be shielded against harassment and coercion, and to obtain restitution for the crime (Fattah, 2000).
15 Concurrent to political changes, self report victimization surveys were initiated and institutionalized in North America in 1966, such as the National Crime Survey. Mass data gathering spurred theoretical formulation and political efforts to enhanc e victim support and rights (Fattah, 2000). Victimization surveys transitioned victimology from small studies of individual crimes to a more macro approach to explaining victimization (Fattah, 2000). The macro perspective successfully informed researchers on the incidence and scope of victimization, but provided little concrete psychosocial information on victim selection and victim -offender exchanges (Fattah, 2000). However, researchers Wesley Skogan, Jim Garofalo, Michael Hindelang, and Michael Gottfredso n published prolifically on victimization surveys and spoke around the nation on their findings, providing further legitimacy to the victims rights movement (Dussich, 2003). Other contributors to the empirical development of victimology and to research on family and interpersonal violence include Marvin Wolfgang (1958), Walter Reckless (1967), and Steven Schafer (1968), Vincent Fontana (1973), Susan Brownmiller (1975), and Del Martin (1978) (Dussich, 2003). Victimology: Philosophies of Science A variety of perspectives comprise philosophies of science in victimology. Mawby and Walklate (1994) divide victimology into three areas: Positive Victimology, Radical Victimology, and Critical Victimology. Positive Victimology emphasizes the use of the sc ientific method to identify typologies of victims and characteristics that precipitate interpersonal criminal victimization (Ben David, 2000). Radical Victimology considers a broader definition of victimization that includes criminal, man made, human right s, and abuse of power victimization (Ben David, 2000). Victims may be considered
16 as entire groups of people, such as workers laboring in a hazardous workplace (Karmen, 2012). Criminal law is viewed as supporting the status quo and rights of the privileged at the expense of addressing poverty, unemployment, and health disparities (Karmen, 2012). According to radical victimologists, laws define and target criminal behavior perpetrated by low power groups (BenDavid, 2000). As a result, although Positive Victi mology focuses on identifying victim proneness, Radical Victimology emphasizes the perpetration of victimization through exploiting structures of power (Ben David, 2000). Critical Victimology primarily examines victim support policies and is informed by Critical Criminology (Ben David, 2000). Karmen (2012) contributes two more philosophies of science: conservative and liberal. Conservative victimologists study street crimes and explore both victim and offender accountability. Personal responsibility is pri oritized over government intervention as a crime prevention strategy Conservative victimology recommends the criminal justice system penalize offenders to seek retribution for victims and deter others committing crimes (Karmen, 2012). Liberal victimologis ts look beyond street crimes to investigate corporate crime and government corruption, advocating governmental intervention to prevent crime (Karmen, 2012). An expanded welfare state including victim compensation, crisis centers, and support programs, is suggested to minimize the consequences of victimization trauma (Karmen, 2012). Restorative justice is one element of liberal victimology, concentrating on victim restitution rather than offender punishment (Karmen, 2012). Critics of radical or liberal stre ams of victimology protest the pain that individual survivors of victimization face is overlooked (Rock, 2007).
17 Models on the Origins of Victimization Models explaining victimization likelihood may be divided into 10 categories: opportunities, risk factors motivated offenders, exposure, associations, dangerous times and dangerous places, dangerous behaviors, high risk activities, defensive/avoidance behaviors, and structural/cultural proneness (Fattah, 2000). The majority of these models consider victim characteristics, activities, and behaviors that increase victim attractiveness or proximity to offenders, such as opportunity, exposure, association, dangerous times and dangerous places, dangerous activities, and high risk activities models. Although moti vated offender models focus on the offender's perspective, these models emphasize the selection rules offenders employ to target victims, similar to models examining victim characteristics, activities, and behaviors (Fattah, 2000). Likewise, although defen sive/avoidance behaviors models consider the converse of dangerous behaviors, these models are still considering victim characteristics, activities, and behaviors that influence victim attractiveness or proximity to offenders (Fattah, 2000). One exception is structural/cultural proneness models, which examine the role of power, stigma, and marginalization in victimization likelihood and assignment of criminal responsibility (Fattah, 2000). With broad similarities among the majority of victimization models the subsequent review of victimization models considers two types : situation -oriented and social -structural/cultural victimization theories. Situation -Oriented Victimization Theories Situation -oriented theories include early victim precipitation models, and view victimization as dynamic and rooted in problematic offender victim interaction (Schneider, 2001). Critics of situationoriented theories view them as victim blaming,
18 although Fattah (1994) argues situation-oriented theories do not assign judgments o r intimate that behaviors associated with victimization should result in assignment of responsibility (Schneider, 2001). Fattah (1991) distinguishes precipitation from provocation, where the former may be used to prevent future victimization, and the latte r is used to attribute responsibility to victims (OConnell, 2008). Schneider (2001) likewise proposes that situation-oriented theories, based off symbolic interaction theory, do not dispute the offender's sole responsibility for the crime, but propose tha t particular victim behaviors are more or less likely to result in an offender choosing to perpetrate a crime. Victim precipitation theories are later echoed in the body of literature through lifestyle theory (Hindelang, Gottredson, & Garofalo, 1978) and r outine activity theory (Cohen & Felson, 1979). Lifestyle and routine activity theories examine crimes that involve physical contact between one or more offenders and a victim of violent crime or theft (Fattah, 2000). The lifestyle theory of victimization p roposes that differences in lifestyle explain much of the variation in victimization likelihood. V ictimization surveys indicate that victims are likely to be male, young, and from a lower socioeconomic status, among other characteristics (Fattah, 2000). Of fenders share similar characteristics, as Hindelang et al. (1978) observed a significant overlap between victim and offender demographics. Patterns of overlapping demographic characteristics of criminals and victims are referred to as propinquity, and pa tterns of nearness between criminal and victim households is called "proximity." V ictims of crime are also more likely to report perpetrating crime than non -victims (Sparks, Genn, & Dodd, 1977). Lifestyle theory transitioned into lifestyle routine activiti es theory, which evaluates not only victim demographics, but also high risk behaviors, such as socializing at night, in situations with
19 alcohol, or with individuals likely to perpetrate crimes (Schneider, 2001). Finally, lifestyle -routine activities theory transitioned into routine activity theory (Schneider, 2001). Routine activities theory studies the impact of daily activities and social context on an individuals vulnerability to victimization. Offenders view a target as appealing based on proximity (i. e. close to motivated offenders), exposure (i.e. in a risky environment), attractiveness (i.e. possess ing valuable property or appearing vulnerable), or level of guardianship (i.e. accompanied by other people or defensive weapons) (Cohen & Felson, 1979). R outine activities theory emphasizes situational over demographic risk factors. Social -Structural and Cultural Victimization Theories In social -structural victimization theory, economic and power structures result in the least powerful members of society ex periencing disproportionate victimization (Schneider, 2001). Schneider (1992) references Australian aborigines as an example of a group of people whose population has been devastated as a result of cultural and institutional discrimination. Feminist crimin ology is one prominent form of social structural victimization theory. Carol Smart critiqued criminology in Women, Crime and Criminology (1977), arguing that women offenders have been marginalized due to criminologists overlooking historical, economic, and political factors influencing crime. From Smarts (1977) perspective, positivist analysis of crime frequency ignores a cultural context where inequality influences involvement in crime. Similarly Susan Brownmiller (1975) and other feminist criminologists considered the role of power in male violence against women, where violence is used as a tactic to coerce women into intimate, economic, and political submission. When Menachem Amir applied the victim
20 precipitation approach to rape in 1967, b acklash by fem inist criminologists significantly undercut the predominance of victim precipitation models in criminology. In the 1980s and 1990s feminist criminology reframed victims as survivors to disentangle the concept of victimization from passivity and submission (Rock, 2007). In addition, feminist criminologists expanded their perspective to consider power in general, where women's experiences were proposed to vary based on class and r ace, rather than gender alone (Bograd, 1999; Crenshaw, 1991; Heise, 1998; Sokoloff & Dupont, 2005). Feminist criminology influenced radical criminology through prompting recognition of the political and social importance of acknowledging interpersonal crime, rather than emphasizing class conflicts alone (Rock, 2007). Lastly, cultural victimization theories examine how societal norms and ideologies result in particular groups being targeted for victimization (Schneider, 2001). Hate crimes are an example of cultural victimization, where members of a group marginalized by societal norms a re targeted such as individuals identifying as or perceived to be lesbian, gay, bisexual, transgender, or intersex (LGBTI). Conclusion Victimology is now an influential sub -discipline of criminology, although the nature of its focus on a stigmatized grou p and its legitimacy as an academic discipline is continually challenged (Fattah, 2000). From victimology's controversial origins in victim precipitation theories, evolution to focusing on structural class victimization, to its ultimate refocus on both int erpersonal and structural crime, philosophies of science and theories of victimization in victimology are diverse and inclusive of many perspectives.
21 Theoretical Background: Models of Sexual Repeat Victimization Presently, the majority of research on repe at sexual victimization studies victim psychology and trauma that act as risk or protective factors against future victimization (Breitenbecher, 2001; Marx, Heidt, & Gold, 2005). In addition, a more limited body of research on repeat victimization likeliho od assesses victim behavioral patterns and situational variables specific to sexual victimization by the initial perpetrator. Psychological and Trauma Factors Psychological/trauma research is the most expansive in the literature on sexual repeat victimization risk Marx et al. (2005) conducted a review of major sexual repeat victimization theories and identified a focus on factors such as attributional style, learned helplessness, chronic physiological hyperarousal, PTSD, negativ e self image, avoidant coping behaviors, poor relational skills, and feelings of powerlessness and stigmatization. Similarly, Breitenbechers (2001) empirical investigation of sexual repeat victimization research categorized the most prominent theories int o general types: disturbed interpersonal relationships, cognitive attributions, self -blame and self -esteem, coping skills, trauma/disassociation, and general psychological adjustment. Distur bed interpersonal relationships, cognitive attributions and self -esteem theories relate intrinsic individual differences to variable repeat victimization risk Of six studies reviewed on dependency insecurity or submissiveness a nd sexual repeat victimization, Breitenbecher, (2001) noted that three documented a signif icant relationship, and three did not (Copeland, 1996; Greene & Navarro, 1998; Knowles, 1993; Leonard, 1992; Mandoki & Burkhart, 1989; Marhoefer Dvorak, Resick, Hutter, & Girelli, 1988) However, the sole prospectively designed study did relate lower
22 asser tiveness and insecurity about cross -gender relationships with increased sexual repeat victimization risk (Greene & Navarro, 1998). Another theory considering problematic intimacy patterns with respect to repeat victimization is dysf unctional interpersonal schema models Cloitre, Cohen, and Scarvalone (2002) documented a relationship between flexibility and interpersonal expectations and repeat victimization The authors surmised that women who learn inflexible relationship schemas through childhood sexual a buse are vulnerable t o forming violent, yet familiar relationships with new abusers in adulthood (Cloitre, Cohen, & Scarvalone, 2002) In contrast to dysfunctional interpersonal schema models, c ognitive attribution theories hypothesize that an internal, stable, and global attributional style relate to sexual repeat victimization risk. Internal, stable, and global attribution styles are evident when survivors attribute responsibility for the victi mization to themselves and believe that the cause of the victimization will inevitably emerge in future situations. These theories suggest that the experience of an uncontrollable traumatic event spur the survivor to view self -protective actions against fu ture threats as fruitless. However, research in female college samples has failed to connect attributional style to the relationship between childhood and adult sexual assault (Mayall & Gold, 1995; Reese -Weber & Smith, 2011). Two studies have documented a significant relationship between self -esteem and repeated experiences of unwanted sexual contact in clinical and college samples (Kellogg & Leonard, 1992; Van Bruggen et al., 2006) with the second study examining sexual self -esteem specifically. Alternatel y another three observed inconclusive findings regarding low self -esteem and repeat contact or noncontact sexual victimization among college and community samples in
23 California and Hong Kong (Chan, 2011; Reese Weber & Smith, 2011; Wyatt et al., 1992). In c ontrast to psychological theories of repeat victimization that propose a static psychological individual difference increases repeat victimization risk, trauma based theories predict that negative psychological outcomes due to sexual victimization result i n increased vulnerability for future victimization. Traumatic bonding theories propose that emotional bonding with an abuser results in continued exposure to the offender and sexual repeat victimization. However, traumatic bonding does not explain sexual r epeat victimization by different perpetrators (Breitenbecher, 2001). Alternately, self -blame models explaining repeat victimization by multiple perpetrators propose self -blame for an initial by one perpetrator diminishes aggressive responses to threats by a new perpetrator. O ne prospective study found both self -blame and decreased sexual assertiveness were associated with repeat sexual assault in a female university sample (Katz et al. 2010) In Filipas and Ullman s (2006) research college women who exper ienced noncontact or contact childhood sexual abuse and subsequent attempted or completed rape reported greater self -blame for the initial sexual abuse at the time of the study than those not revictimized in adulthood Research on coping strategies proposes that poor coping strategies in response to an initial sexual assault, such as avoidance, substance use, and isolation, compound vulnerability to repeat victimization. However, case studies have not documented differe nt coping strategies between repeat victims of contact or noncontact childhood sexual abuse and adult sexual assault and victims of childhood sexual abuse only (Mayall & Gold, 1995; Proulx Koverola, Fedorowicz, & Kral, 1995), or identified maladaptive
24 cop ing strategies as predictive of experiencing both childhood sexual abuse and repeat attempted or completed rape (Najdowski & Ullman, 2011; Filipas & Ullman, 2006). Notably, the two studies observing a relationship operationalized adult sexual victimization as rape only, rather than sexual assault. Considering more closely the sole study examining a non-college sample, Najdowski and Ullman (2011) found differences in coping styles after the initial victimization between women who were revictimized and those that were not. Specifically, maladaptive coping, or disengagement, denial, humor, self -blame, self -distraction, and substance use at Time 1 was significantly associated with repeat sexual assault one year later (Najdowski & Ullman, 2011). Social withdrawal is characteristic of maladaptive coping. In a model that included severity of first victimization, PTSD, and coping strategies, only maladaptive coping strategies (substance use, social isolation, and acting out sexually or aggressively) predicted sexual repeat victimization (Filipas & Ullman, 2006) Another recent study documented avoidant coping as predictive of increased trauma after contact childhood sexual abuse, and greater trauma symptoms as predictive of higher rates of later sexual assault among a female college sample (Fortier, DiLillo, Messman Moore, Peugh, DeNardi, & Gaffey, 2009). However, research finds that simply talking with others about a sexual victimization does not differentiate repeat victims of noncontact or contact childhood sexual abuse and rape in adulthood from those who experienced childhood sexual abuse only (Filipas & Ullman, 2006). Poor threat perception is a frequently studied trauma symptom with respect to sexual repeat victimization, where threat perception is presum ed to be compromise d as a result of initial victimization Threat perception has been operationalized in several
25 studies as accuracy in perceiving sexual interest and response latency when indicating that a man has gone too far in a hypothetical aggressi ve hetero sexual encounter Research in female college samples finds mixed results in the relationship between perception of threat and repeat sexual assault or rape experiences (Kearns & Calhoun, 2010; Marx, Calhoun, Wilson, & Meyerson, 2001; Messman-Moore & Brown, 2006). Inconclusive findings may be due to threat perception indirectly influencing sexual repeat victimization. Kearns and Calhoun (2010) find behavioral aggression in response to threat more consistently relates to repeat rape victimization tha n threat perception. Furthermore, research documents college women's self -blame for an experience of unwanted sexual contact or sexual advances as a mediator between threat appraisal and diminished aggression (Katz et al. 2010; Nurius, Norris, Young, Graha m, & Gaylord, 2000). Among studies assessing PTSD as a risk factor for repeat sexual victimization in college samples some research finds symptoms to be a risk factor for repeat contact sexual victimization (Messman Moore, Ward, & Brown, 2009; Walsh, DiL illo, Klanecky, & McChargue, in press ; Sandberg Matorin, & Lynn, 1999), whereas other research does not link PTSD to adult sexual assault among respondents reporting experiencing "something sexual" in childhood (Arata, 1999), or suggests arousal PTSD symptoms may enhance recognition of sexual threats (Wilson, Calhoun, & Bernat 1999). Mixed findings on PTSD as a risk factor for sexual repeat victimization may be due to mediators, as Messman -Moore et al. (2009) observed that substance use mediated the re lationship between PTSD symptomology contact childhood sexual abuse, and rape victimization in college In a similar study, Ullman, Najdowski, and Filipas (2009)
26 assessed the relationships among childhood sexual abuse, adult sexual assault, PTSD symptoms, and substance use in a sample of women 18 or older Participants responded to advertisements to join a research study on experiences of sexual assault since the age of 14, or unwanted sexual contact, sexual coercion, completed or attempted rape PTSD was measured through four symptom clusters: re -experiencing, avoidance, arousal, and numbing. Re -experiencing symptoms comprise upsetting thoughts and nightmares or the emotional or physiological feeling that the trauma is re-occurring Avoidance symptoms invo lve deflecting thoughts about the experience, and arousal includes feeling overly alert or irritable. Lastly, numbing symptoms manifest as an inability to remember the victimization, diminished interest in activities or socializing, and lack of emotion. Nu mbing symptoms were found to mediate directly the relationship between childhood sexual assault and sexual repeat victimization at Time 2. However, reexperiencing, avoidance, and arousal symptoms were not directly related to re peat victimization, although these three PTSD symptoms did predict problem drinking, which then predicted sexual assault at Time 2 (Ullman et al., 2009) Lastly, research on general psychological adjustment theories studies have connected poor psychological adjustment, depression, or anxiety with experiences of contact or noncontact childhood sexual abuse and repeat adult sexual assault in two prospective university-sampled studies (Gidycz Coble, Latham, & Layman, 1993; Greene & Navarro, 1998). Yet Gidycz et al. (1995) failed in a replication to find an effect of psychological adjustment on the relationship between experiencing contact or noncontact childhood sexual abuse and sexual assault in college. A more recent longitudinal study on risk factors for repeat peer -to -peer physical violence, childhood
27 maltreatment, and sexual violence additionally documented a relationship between psychological distress and repeat victimization (Cuevas, Finkelhor, Clifford, Ormrod, & Turner, 2010). Findings in a community sample indicated that that general well -being did not relate to women's repeat experiences with nonconsensual contact or noncontact sexual experiences (Wyatt, et al., 1992) Behavioral Factors Interpersonal behaviors that are framed as "unhealthy" or "risky" are commonly exami ned in repeat victimization research. Alcohol use is one example, and two prospective examinations following cohorts over time find alcohol use predicting repeat contact sexual victimization (Greene & Navarro, 1998; Testa, Hoffman, & Livingston, 2010). An other frequently examined risk factor with respect to repeat victimization is sexual behavior. S everal studies using college, clinical, and European community samples have associated frequent consensual sexual activity with repeat contact sexual victimization ( Fargo, 2009; Himelein, 1995; Krah, Scheinberger Olwig, Waizenheffer, & Kolpin, 1999) Other research in college, military, and African -American clinical samples have found a minimal or non -significant relationship between sexual behavior a nd sexual repeat victimization ( Collins, 1998; Gidycz et al., 1995; Mayall & Gold, 1995; Merrill et al. 1999; West et al., 2000). It should be noted that Mayall and Gold (1995) and West et al.s (2000) research used a broader definition of sexual victimiz ation, encompassing both contact and noncontact sexual assault and "unwanted sexual experiences" respectively. Alternately, Fargo (2009) assessed nonconsensual genital contact with someone five or more years older than the respondent when they were 17 year s old or younger and nonconsensual sexual contact when the respondents were 18
28 years or older. More specifically, Fargo's (2009) research assessed longitudinal data collected over 24 years on a range of risky behaviors, including age s at first alcohol use, drug use, and consensual sex. In addition, Fargo (2009) measured whether the respondent had run away from home as a child or adolescent, total number of consensual sex partners, whether the respondent had been involved in prostitution, and frequency of al cohol consumption. Fargo (2009) found that any r isky behavior in adolescence mediated a relationship between child and adolescent sexual assault, and risky sexual behavior mediated the relationship between adolescent and adult sexual victimization. Likewis e, risky sexual behavior mediated the relationship between adolescent risky behavior and repeat adult sexual victimization. Lastly, frequent alcohol use before sex and in general indirectly related to adolescent risk -taking beh avior and risky sexual behavi or ( Fargo, 2009). Situational Factors Although some sexual repeat victimization studies consider situational or environmental factors, Breitenbecher (2001) considers this research to be the exception to the rule in a field where victim psychology perspectives prevail Since Breitenbecher (2001) conducted his review, researchers ha ve begun exploring risk factors emerging from the context of victimization. Gr iffin and Read (2012) document that experiencing physical force as a coercion tactic during c hildhood or adolescent sexual assault resul ts in higher adult sexual assault risk compared with those who experienced incapacitation by drugs or alcohol as a method of coercion. A recent sexual assault general population survey contrasting singly victimize d women, women victimized multiple times by one
29 perpetrator and women victimized multiple times by distinct perpetrators found that more severe initial sexual assaults and younger age at the time of first assault were associated with both ongoing and repeat victimization (Casey & Nurius, 2005). Women repeat sexually assaulted by different offenders additionally had greater non -sexual trauma in childhood (Casey & Nurius, 2005). West et al. (2000) examined a sample of Black women with histories of child hood sexual abuse, or sexual contact by force, threat of force, coercion, authority, or someone who was five or more years older, documented in emergency room records An average of 17 years after initial participation in the research, 136 women were locat ed and agreed to participate in re -interviews. Sexual re peat victimization was considered to be unwanted sexual experiences, sexual assault, or rape after the age of 18. Respondents were asked whether the perpetrator of childhood sexual abuse hit, pushed, beat, slapped, choked, gagged, or used a weapon to force them into sexual contact. In addition, the participants were asked whether oral, anal, or vaginal penetration was involved, or whether the abuser was intra -familial or extra -familial. Re peat victimiz ation was associated with physical violence as a coercion tactic but penetration and abuse by a family member did not predict sexual repeat victimization (West et al., 2000) When assessing sexual offender coercion tactics as a risk factor for future sexu al victimization, it is important to consider a relevant covariate: victim -offender relationship Research finds curvilinear relationships between victim -offender relationship and use of violence, where either stranger and intimate offenders use greater vi olence than acquaintances (Mller, Bckstrm, Sngergaard, & Helstrm, in press ; Stermac Del Bove, & Addison, 2001; Ullman & Siegel, 1993), or intimates cause greater
30 injury compared with strangers and acquaintances (Murphy, Potter, Pierce -Weeks, Stapleto n & Wiesen -Martin, 2011). Although less research has been done on police involvement and repeat sexual victimization, Jordan (2004) found arrests and protective orders to be moderately effective in lowering repeat victimization rates among victims of partn er violence. Summary Breitenbecher (2001) concludes that although individual level psychological theories have intuitive appeal, they have received little empirical support (pp. 428), with the exception of alcohol use and threat perception research A n empirical emphasis on victim psychology and behavior is a natural byproduct of information access survivors of sexual victimization are naturally more forthcoming sources of information than perpetrators. As Miller, Handley, Markman, and Miller (2010) n ote victimization information often frames perpetrator behavior in a manner that suggests they are the background of the story. As a result, counterfactual s emphasize changes in victim behavior to avoid the victimization, with the assumption that perpet rator behaviors are a constant. Accordingly scholarly suggestions to expand the sexual repeat victimization focus beyond individual factors may not be a critique of victim self report data, so much as a recommendation to frame victim focused risk factors such as alcohol use or frequent sexual encounters within a larger social context Sociocultural environments assign or withhold perpetrator responsibility for particular types of sexual assault, such as assaults where victims willingly imbibe alcohol or ha ve a sexual history with the perpetrator. Reframing victim risk factors within a theoretically cohesive framework that considers
31 social and cultural contexts requires a comprehensive consideration of psychological, situational, and sociocultural factors.
32 CHAPTER III POST SEXUAL VICTIMIZATION MENTAL HEALTH Theoretical Background: Models of Post -Sexual Victimization Mental Health The application of PTSD to the range of symptoms survivors of sexual assault experience granted academic and medical legitimacy to post -sexual victimization trauma (Goodman, Koss, & Russo 1993; Herman, 1992) However, critics of a post -sexual assault PTSD perspective argue the model overlooks the soci o cultural context that sexual assault s occur within (Campbell Dworkin, & Cabral 2009). Proponents of expanding sexual assault recovery theories suggest that post -rape trauma originate s not only in the assault itself, but in cultural responses to this crime, its victims, and its perpetrators. Alternative perspectives include socio -ec ological frameworks, or those that assess individual and extra individual origins of psychological distress and recovery post -sexual assault (Campbell et al., 2009; Harvey, 1996) Ecological frameworks, also known as human ecology models consider relation ships among groups of people within particular physical and social environments, reminiscent of biologists study of organisms relat ionships to their surrounding environment. Several scholars have proposed ecological models of post -sexual assault distress that situate post assault trauma within sociocultural responses to sexual victimization (Charuvastra & Cloitre, 2008, Neville et al. 2004, & Samuels Dennis et al., 2010). Suggestions to include socio-cultural factors in models of post -sexual assault recovery have been heeded. A wide range of risk factors are proposed to explain the relationship between sexual victimization and psychological trauma, including variables perta ining to victim characteristics, the context of the sexual assault itself, and the post -
33 assault experience. The majority of research on risk factors for post -sexual victimization trauma does not conduct differential analyses based on victimization status ( victimization by one perpetrator versus multiple victimizations by separate perpetrators ). However, post victimization distress may heighten vulnerability to future victimization (Gidycz et al., 1993; Greene & Navarro, 1998) and d ifferential experiences i n the victimization and post victimization context may trigger unique mental health trajectories influencing whether one may experience future sexual assaults by new perpetrators Although few studies on post -victimization distress differentiate individua ls victimized by one perpetrator from repeat victims, the literature on risk and protective factors for post victimization recovery still contributes a breadth of information on victim, sexual assault, and post assault variables relating to post assault re covery. The current study will contribute to the literature on post -sexual assault distress through extending the search for risk factors by specifying separate multivariate models for those sexually victimized by one perpetrator only and repeat victims. Victim Characteristics Vi ctim characteristics such as race/ethnicity, age, education, employment, pre assault mental health diagnoses, and victimization history are assessed across multiple studies on post -sexual assault distress Examinations o f responses to sexual assault or rape with respect to r ace or ethnicity have largely found no differences in PTSD symptoms in community samples (Campbell et al., 1999; Elliott, Mok, & Briere, 2004; Masho & Ahmed, 2007; Ullman et al. 2006), depression symptoms in community and clinical samples (Ahrens, Abeling, Ahmad, & Hinman, 2010; Elliott et al., 2004; Frank &
34 Stewart, 1984; Sorensen & Siegel, 1992), and suicidal ideation/attempts among college and community sampled responde nts (Brener, McMahon, Warren, & Douglas, 1999; Ullman & Brecklin, 2002b). However, some research finds Hispanic/ Latina women reporting greater PTSD or anxiety after a rape than White non Hispanic/Latina or African American women (Littleton, Grills Taquechel, Buck, Rosman, & Dodd, in pres s ; McFarlane et al., 2005 ), and White survivors of rape more likely to report rece nt substance use than women and girls of color (Littleton et al., in press ; Resnick et al. 2013). In Sciolla, Glover, Loeb, Zhang, Myers, and Wyatt s (2011) sample, attempted or completed childhood rape predicted symptoms of depression for Latina, but not African American women. The influence of race and ethnicity on post assault recovery may be mediated through self blame, with research suggesting that women of color experience greater cultural attributions of blame for their sexual assault s than White women (Neville et al. 2004; George & Martnez, 2002). Generally research on income and employment status has not found a relationship to post -sexual assault distress in community and national samples (Ullman & Brecklin, 2002a; Ullman & Filipas, 2001b) However, a recent study on diverse traumas, including assaultive traumas, determined personal resources mediated the relationship between a traumatic experience and PTSD (Samuels Dennis et al., 2010) Bryant Davis, Ullman, Tsong, Tillman, and Smith (2010) corroborated Samuels Dennis et al.s (2010) findings in their examination of PTSD, depression, and illicit drug use among African American women who experienced unwanted sexual contact attempted rape, or completed rape Sexual victimization prior the age of 14 was considered child sexual abuse, and sexual victimization after the age of 14 was referred to as sexual assault PTSD symptoms were
35 identified through reports of re -experiencing, a voidance, arousal, and numbing and remaining measures assessed suicidal ideation, levels of depression, and illicit drug use in the prior year. When controlling for a history of sexual abuse in childhood, women experiencing repeat sexual assault in adulthood with lower incomes reported more depression and PTSD symptoms than those with higher incomes. The odds of lower income repeat sexual assault victims reporting illicit drug use were 18% higher than those of repeat sexual v ictims with higher income s (Bryant et al., 2010) A related construct to income, education level, is found in an equivalent number of studies to have no influence (Campbell et al., 1999; Ullman et al., 2006; Ullman, Townsend, Filipas, & Starzynski, 2007), or an ameliorating effect on the psychological aftermath of a sexual assault or rape for certain subgroups in community samples (Ullman & Brecklin, 2002a; Ullman & Filipas, 2001b). More recent research o n a random -digit -dialing sample in Virginia connected income below $49,000 with greater odds of PTSD symptoms than the comparison group of $75,000 (Masho & Ahmed, 2007). A n education level of less than high school in comparison to postgraduate education additionally related to greater odds of PTSD among cont act child sexual abuse and adult rape survivors (Masho & Ahmed, 2007). Lastly, the experience of repeat sexual victimization is associated with more severe psychological and interpersonal outcomes than sexual victimization by one perpetrator only (Arata, 1999, 2000; Banyard et al., 2001; Classen et al., 2001; Kimerling et al. 2007; Messman-Moore, Long & Siegfried, 2000). In a national probabil ity sample of 15 to 55-year -old girls and women in the continental United States, respondents sexually molested or raped in both childhood and adulthood were around three times
36 more likely to attempt suicide than those who were not repeat victimized (Ullman & Brecklin, 2002b) Resnick et al. (2013) investigated the influence of repeat violent victimization history on substan ce use among girls and women receiving sexual assault medical care at a medical center. Violent victimization history comprised experiences with rape physical assault committed with the intent to injure or kill the respondent, or assaults with a we apon or that caused an injury. Findings revealed recent or previous marijuana or illicit drug use was associated with repeat assault victimization (Resnick et al., 2013) For women with a history of repeat victimizations and substance misuse, the first assault generally preceded the initiation of substance use (Resnick et al., 2013) Other predictors were not separately modeled based on repeat victimization history Context of the Sexual Assault Vari ables relating to the context of the sexual assault to post assault distress include victim perpetrator relationship, degree of perpetrator violence and victim injury, and substance use in predicting post -sexual assault trauma. Studies on convenience commu nity samples indicate no differential level of trauma based on victim -perpetrator relationship (Ullman & Filipas, 2001b ), a marginal relationship between stranger sexual assault and greater post assault distress (Ullman et al., 2006), or a significant rela tionship between sexual assault by partners or known offenders and greater PTSD or depressive symptomology (Demaris & Kaukinen, 2005; Masho & Ahmed, 2007). In Temple, Weston, Rodriguez and Marshall s (2007) sample of low income women, sexual assault perpetration by a current intimate partner predicted PTSD more strongly than assault by a former partner or non -partner Similarly, Culbertson and Dehle (2001) determined that
37 sexual assault by a spouse, cohabitating partner, or acquaintance correlated wit h more hyperarousal symptoms among college women than sexual assault committed by a date or current intimate partner Further, sexual assault by a spouse or livein partner predicted more intrusive symptoms than assault committed by other perpetrator types (Culbertson & Dehle, 2001) Finnish university students in a recent study by Bjrklund, HkknenNyholm, Huttunen, and Kunttu (2010) reported a range of physical and mental health symptoms that varied based on their relationship to perpetrators of physica l or sexual assault. Parent offenders predicted general and mental health symptoms, intimate current and ex -partner offenders predicted abdominal symptoms, and stranger offenders predicted alcohol consumption (Bjrklund et al., 2010) Whereas earlier research examining the role of violent threats or weapons on post sexual assault distress i n criminal justice, rape crisis center and community samples found little relationship (Campbell et al., 1999; Frank Turner, & Stewart, 1980; Kilpatrick Veronen, & B est 1984), more recent research has associated PTSD with victims who were assaulted with a weapon (Bownes OGorman, & Sayers, 1991), or who had a greater perceived life threat during the sexual assault (Masho & Ahmed, 2007; Ullman & Filipas, 2001b). Likewi se the use of physical violence as a sexual assault coercion tactic has been linked with post assault distress in some college and community samples (Brown, Testa, & Messman -Moore, 2009; Littleton & Henderson, 2009; Ullman & Siegel, 1993), or marginally or unrelated to depression, PTSD, substance use, and anxiety in other assessments of university and rape crisis center respondents (Atkeson, Calhoun, Resick, & Ellis, 1982; Griffin & Read, 2012; Kilpatrick et al., 1984). Regarding infliction of an injury beyond the sexual assault or rape itself, two studies using national
38 or random -digit -dialing samples found a positive relationship with post assault distress (Acierno Resnick, Kilpatrick, Saunders, & Best, 1999; Resnick, Kilpatrick, Dansky, Sau nders, & Best, 1993) and a separate study on college students found no relationship (Ullman & Filipas, 2001b). The relationship between coercion tactics and post assault distress may not be entirely straightforward. Brown, Testa, and Messman -Moore (2009) d ocumented respondents reporting that trauma levels due to alcohol/drug incapacitated rape fell between those due to physically forcible rape and verbal coercion, as anticipated. However, perpetrator attributions of responsibility and scores on particular d omains of trauma (current levels and emotional impact) were equivalent for incapacitated rape and forcible rape (Brown et al., 2009). Another study reveals that lower levels of physically violent coercion tactics were associated with being an unacknowledge d victim, and greater physical violence was indirectly related to PTSD symptomology through maladaptive coping in college women (Littleton & Henderson, 2009). Early r esearch on substance use prior to an assault found no connection to post rape or sexual as sault distress in community samples (Campbell et al., 1999; Resnick, Yehuda, & Acierno, 1997) In contrast, Griffin and Read (2012) related incapacitated sexual victimization among college women to higher scores on problem drinking measures Likewise, Resn ick et al. s (2013 ) survey of sexual assault medical services patients found survivors whose sexual assaults involved substance use indicat ed higher rates of alcohol use and misuse 6 months later. In a national phone survey, Zinzow et al. (2010) linked experiences with substance -facilitated rape and/or rape by force or threat of force to PTSD, and forcible assault to depression. Victim substance use and lower levels
39 of physically violent coercion in sexual assault or rape have also been associated with greater victim self -blame and subsequent distress in samp les of college students and low income women (Brown et al. 2009; Tucker, Wenzel, Straus, Ryan, & Golinelli, 2005), although Miller et al. (2010) found no relationship between university women' s alcohol use during a sexual assault and subsequent self -blame. Post -Sexual Assault Experience Research on p ost -sexual assault trauma ha s extensively investigated victim social support and coping styles. P ositive informal or formal social support has been found to alleviate post rape or sexual assault distress in college and community samples (Campbell, Ahrens, Sefl, Wasco, & Barnes, 2001; Orchowski, Untied, & Gidycz, in press ), or have little impact in one London criminal justice sample (Andrews, Brewin, & Rose, 2003) Although these studies findings conflict, i t should be noted that Andrews et al. s (2003) measured social support differently than Campbell et al. (2001) and Orchowski et al. ( in press ). Andrews et al. (2012) measured whether discl osures of sexual victimization received emotional support, practical support, or negative responses. Alternately, the other two studies implemented versions of Ullmans (1996b 2000) Social Reactions Questionnaire (SRQ) describing between 12 and 48 social reactions to a disclosure Campbell et al. (2001) and Orchowski et al.s (in press) findings significantly associating positive social support with post assault recovery may indicate that the SRQ is a more sensitive measure of the range of disclosure responses than that used in Andrews et al. (2003) Another measure, the Social Support Questionnaire (Sarason, Sara son, Shearin, & Pierce, 1987), detected that high quality social support (someone the
40 respondent can count on and accepts the flaws of the resp ondent) wa s a protective moderated mediator for the relationship between severe sexual trauma and problematic substance use, whereas higher quantity of social support wa s a risk moderated mediator, potentially due to heightened alcohol use through greater socializing (Johnson & Johnson, in press ). Negative reactions to sexual assault disclosure are more consistently connected to increased psychological distress and/or al cohol use among sub-populations (Borja, Callahan, & Long, 2006; Campbell et al., 2001; Jacques Tiura Tkatch, Abbey, & Wegner, 2010; Ullman & Filipas, 2001b ; Ullman et al. 2007; Ullman, Starzynski, Long, Mason, & Long, 2008) to an extent greater than the beneficial impact of positive social reactions (Andrews et al., 2003; Davis, Brickman, & Baker, 1991; Campbell et al., 2001; Ullman, 2010). Negative responses to initial disclosures may be the most consequential, at times silencing victims for years (Ahre ns, 2006). Survivors of childhood sexual abuse face unique challenges with disclosures. Among victims of childhood sexual abuse, delayed disclosure, negative reactions to childhood disclosures, and self -blame at time of abuse for survivors of sexual abuse by relatives are associated with greater PTSD symptomology (Ullman, 2007). Other elements of sexual assault disclosures associated with negative reactions are assaults involving alcohol use (Ullman & Filipas, 2001a), and assault committed by relatives or s trangers in comparison to acquaintances or romantic partners (Ullman, 2007; Ullman et al., 2006). Negative reactions based on victim offender relationship were found in both studies to predict PTSD symptoms (Ullman, 2007; Ullman et al. 2006). In addition to PTSD symptomology, n egative social support has been documented in convenience community samples as predicting greater
41 "avoidance coping," where victims socially withdraw, stayat home, and use substances (Ullman et al. 2007; Ullman, 1996c ). Similar to negative social reactions, avoidance coping hinders recovery from sexual assault (Gutner, Rizvi, Monson, & Resick, 2006; Frazier, Mortensen & Steward, 2005; Merrill, Thomsen, Sinclair, Gold, Milner, 2001; Ullman, et al., 2007; Ullman, Relyea, Peter Hagene & Vasquez, 2013; Najdowski & Ullman, 2011). Najdowski and Ullman (2011) examined coping strategies and depression symptoms in a Chicago convenience community sample of women reporting one or more unwanted sexual experiences, ranging from unwanted sexual contact to rape R epeat victims were more likely to report depression than victims assaulted in childhood only. A path analysis of data collected at two points over a year, Time 1 and Time 2, revealed that re peat sexual victimization directly related to de pression at Time 2 M aladaptive coping at Time 1 in response to childhood sexual abuse predicted repeat sexual victimization Maladaptive coping at Time 1 also indirectly related to higher depression scores at Time 2, through predicting maladaptive coping at Time 2, in response to the repeat sexual victimization (Najdowski & Ullman, 2011) These findings suggest that coping strategies in response to childhood sexual victimization link directly with repeat victimization and indirectly with long -term depressi on. Alternately adaptive strategies, consisting of communicating feelings and pursuing social support are generally associated with speedier and more successful recoveries (Frazier, Tashiro, Berman, Steger, Long, 2004; Frazier, et. al., 2005; Gutner et al ., 2006; Runtz & Schallow, 1997; Steel, Sanna, Hammond, Whipple, & Cross, 2004). However, some research complicates the relationship between adaptive strategies and
42 post sexual assault recovery. Brand and Alexander (2003) assessed a mid -Atlantic newspaper recruited sample of childhood sexual abuse survivors finding both avoidance and seeking social support predicted greater adult dysfunction in work, social life, extended family, intimate relationships, and financial management, whereas distancing predicte d less dysfunction. Johnson Shehan, and Chard s (2003) clinical sample of contact childhood sexual abuse victims connected approach coping strategies with passive aggressive and histrionic personality disorder, and avoidance coping strategies with PTSD sym ptomology and avoidant, dependent, borderline, paranoid, schizotypal and schizoid personality disorders. Greater access to friends, relatives, and religious services protected against PTSD and depression in a convenience Chicago sample of African -American sexual assault survivors, although r eligious coping predicted greater levels of depression and PTSD symptoms (Bryant -Davis, Ullman, Tsong, & Gobin, 2011). The formal alternative to victim support from friends, family, and partners includes the criminal justice and medical care system s Rates of survivors seeking formal support range from 18% to 40% of survivors filing a police report and seek ing prosecution, and 9% to 60% of survivors speaking with mental health professionals (Busch -Armendariz, Bell, DiN itto, & Neff, 2003; Campbell et al., 1999; Ullman, 1996a; Ullman & Filipas, 2001a; Sedgwick, 2006, Kilpatrick, Amstadter, Resnick, & Ruggiero, 2007). Sexual assault v ictims are more likely to seek formal support when strangers perpetrate crimes (78% vs. 57.6%), when the re are injuries beyond the sexual assault itself (73.2% vs. 55.7%), and when perpetrators threaten victims lives (71.6% vs. 55.1%) (Ullman & Filipas, 2001a). However, research has indicated formal support systems may
43 be associated with probl ematic responses to victim disclosures. Negative social reactions (controlling, egocentric) are more common for women telling formal than informal support sources, particularly for women of color regarding egocentric reactions (Ullman & Filipas, 2001a). Se xual assault v ictims in a Chicago convenience sample reported more negative reactions from formal support sources than i nformal support sources and more victim blaming, controlling, stigmatizing and egocentric responses when assaults involved alcohol use (Ullman & Filipas, 2001a). Expectations of negative reactions by formal support systems influence willingness to report a sexual victimization. One clinical study found women who did not report their sexual assault were more likely to believe that police w ould be insensitive or victim -blaming compared with women who did report their sexual assault (Jones, Alexander, Wynn, Rossman, & Dunnuck, 2009). Secondary victimization, where post assault services are delivered insensitively or accusably can substantially interfere with recovery. Campbell and Raja (2005) surveyed sexual assault survivors at a Veterans Administration Hospital Womens Clinic regarding their experiences with legal and/or medical systems post assault. Fifty nine percent of women were encouraged by legal and/or medical personnel not to report the assault 65% were asked about their sexual history, an d between 65% and 82% of women felt guilty, depressed, anxious, distrustful, and rel uctant to seek further support consistent with ot her findings (Campbell, 2005). Victims whose assaults do not conform to rape scripts, such as acquaintance rape, report more victim -blaming questions, and women with lower SES, and women of color describe facing greater challenges seeking assistance (Campb ell, 2008). Research indicates that experiences with secondary victimization and unwillingness by the criminal justice system to pursue rape prosecution
44 is associated with more severe PTSD (Campbell et al., 1999; Campbell & Raja, 2005), although Walsh and Bruce (2011) found that the further a rape prosecution went in the legal system, the less perceived control respondent s reported regarding their current distress and recovery. Masho and Ahmeds (2007) random -digit -dialing sample of adult rape or contact childhood sexual abuse survivors indicated that respondents recounting talking to a crisis hotline or a counselor displayed greater PTSD symptomology However, PTSD symptoms may reflect a more traumatic assault, spurring survivors to reach out to multiple fo rms of social support. Alternately, talking with police was associated with less PTSD (Masho & Ahmed, 2007). Summary Although post -sexual assault mental health and coping strategies certainly reflect genetic predisposition s and post traumatic stress research determining negative post assault experiences hinder recovery provide s support for the perspective that post victimization distress does not occur in isolation to culture Post -sexual assault distress emerges not only from the severity of victimization and victim coping strategies, but additionally from societal response s to the victim and assignment of responsibility. In spite of repeat sexual victimization broadly documented as a risk factor for post victimization distress, unique post assault re covery trajectories of those victimized by one perpetrator only versus multiple victimizations by separate perpetrators have received rare consideration The present study considers whether variations in individual factors, first sexual victimization conte xt, and post -victimization experiences distinguish longterm mental health outcomes between respondents sexually victimized by one offender only and repeat victims.
45 CHAPTER IV ECOLOGICAL FRAMEWORK A major shortcoming to the body of literature on sexual repeat victimization risk is the lack of organizational structure categorizing predictors of repeat victimization and post -sexual victimization mental health. Formulating explanatory theor ies of sexual repeat victimization and post -sexual victimization men tal health is challenging without an inclusive, empirically driven framework organizing risk and protective factors. Likewise, the construction of an effective policy response to this complex social problem necessitates an understanding how individual, situational, and socio cultural level factors contribute to vulnerability to sexual repeat victimization and trauma. From a normative perspective, the heavy focus o n individual risk factors is problematic, as most available research encourages a victim -b lame interpretation (Grauerholz, 2000, pp. 5). A consequence of a victim -centered research focus is policy recommendations targeting potential victims rather than broad, institutional change. Recommendations urging women to not socialize drink alcohol or interact with men restricts women's access to society, and place s the burden of crime prevention on likely victims, rather than the criminal justice system (Day, 1994). Likewise, critics of a PTSD post victimization mental health perspective protest that policy implications would emphasize the victimization alone overlooking the role of cultural responses to the crime in victim resilience and recovery. Due to the mixed findings of extant repeat victimization likelihood studies (Breitenbecher, 2001), and the victim focused implications of individual level theories, the current study employs an ecological f ramework, combining individual factors with
46 situational, and sociocultural variables (Belsky, 1980; Brofenbrenner, 1977; 1979). Further, the present study continues the line of ecological investigation of post -sexual assault mental health (Charuvastra & Cloitre, 2008, Neville et al. 2004, & Samuels Dennis et al., 2010) with an application to sexual repeat victimization. E cological framework s o f sexual repeat victimization likelihood and trauma avoid exclusive emphasis on individual factors and integrate individual situational, and sociocultural factors to potentially construct a better specified model than those of individual level theories o nly. An ecological model of repeat victimization not only consolidates variable concepts but approaches them from a particular perspective Th e ecological perspective considers how offenders exploit victim history, victim behaviors, and social settings to perpetrate crimes that are unlikely to result in formal or informal sanctions. Moreover an ecological perspective considers cultural standards dictating which victims and situations are likely to generate societal response s critical of the victim and forgiving of the perpetrator Similarly an ecological theory of post -victimization mental health proposes trauma not only originates in the sexual assault itself but in cultural treatment of the crime, victims, and perpetrators. T he ecological model exam ines three nested hierarchical groupings of variables : 1) childhood development, 2) context of the sexual victimization and 3) social power (See Figure 4 1). The first two categories describe how one's personal history and the context of the sexual victim ization relate to perpetrator aggression. S ocial power variables include formal and informal social organizations, such as ones job, community, social support system, and the distribution of wealth (Grauerholz, 2000). The term "social power" is based on t he conceptualization of power as the ability of an individual to
47 influence o thers to their will, commonly for goals of security or advantage (Cartwright & Zander, 1968; Weber, 1998). This definition of power encompasses social identities, such as race or e thnicity, as group level indicators of sociocultural influence. Figure 4.1: Conceptual Groupings of Variables in the Ecological Framework Framework Assumptions T heories explaining violence against women originated through bringing a gendered power disparity as enabling a certain group, men to commit violence against another group, women More recent theories of violence against women have expanded to focus on the intersectionality of multiple oppressive hierarchies, where partner and sexual violenc e is understood to not only impact women differently than men, but to be experienced differently by unique subgroups of men and women, based on other identities linked to variation s in power, such as class, race, and sexual orientation (Bograd, 1999; Crens haw, 1991; Heise, 1998; Sokoloff & Dupont, 2005). Physical and sexual violence committed against women and children by known perpetrators was initially regarded as a private, family dysfunction, then a psychological issue of perpetrator deviance, and until recently as a component of a large r system of power that differently impacts women and children versus men (Crenshaw, 1991). However, a focus
48 on men as perpetrators and women and children as victims overlooks intra -group differences influencing each woman's experienced reality, such as race and class, in addition to gender. I ntersectional approach es avoid the essentialization of the concept of "woman", but are not purely postmodern Simply because identity categories are socially constructed they are not considered to be irrelevant to patterns such as pay or health disparities (Crenshaw, 1991). For example, queer and heterosexual may each comprise ends of an envisioned continuum of sexuality, but one, heterosexuality, is legally and culturally privileged o ver the other -as witnessed in marriage inequality worker protection rights inconsistently protecting gays and lesbians from termination due to their sexualities, and culturally, through hate crimes and media representations (Bograd, 1999). How a social identity such as sexuality intersects with other dimensions such as race and socioeconomic status influences how victimization is experienced and responded to, how personal versus social responsibilities are portrayed, and the degree to which a party is saf e from victimization (Bograd, 1999). Context can also vary patterns of social power (Bograd, 1999). Students who attend traditional four year colleges are on average from wealthier backgrounds than the general population and consequently may be presumed to experience preferential criminal justice treatment. However the risk of sexual assault on college campuses is considerable, with research documenting that among survivors of sexual violence, female college students are more likely to e xperience a new sexual assault than non -students (Kilpatrick et al., 1998). Some university systems have been decried for their response to campus sexual assault commonly involving pressuring victims to participate in university administrative proceedings that lead to either academic penalties or no
49 repercussions for their alleged assailants (Lombardi, 2009). As a result, financial or gender inequality alone cannot account for the origins of sexual violence. Unilateral theories on gender inequality may exp lain why women are so frequently victimized by male intimates, but they do not clarify why most men do not sexually assault women, or why all women are not sexually assaulted (Heise, 1998). An ecological framework differentiates why one victim may be targe ted in one situation or moment in time and not another (Heise, 1998), or why recovery from a victimization may vary within and between social identities. Ecological Risk and Protective Factors for Sexual Repeat Victimization and Trauma In order to further specify how each grouping of variables in the ecological framework applies to the concepts of sexual repeat victimization likelihood and trauma, risk factors are discussed by each ecological category : childhood development context of the first victimizat ion and social power Childhood D evelopment Sexual Repeat Victimization A variety of variables that have been linked to sexual repeat victimization, may be categorized as childhood development risk factors, or factors that comprise an individuals ba ckground or personal history. Childhood physical abuse is one example of a personal history variable that pertains to sexual repeat victimization likelihood, and research documents its relationship to adult physical and/or sexual victimization for both men and women (Desai et al. 2002). Early parenthood has largely been examined as a consequence of rather than as a risk factor for repeat victimization (Friesen, Woodward, Horwood, & Fergusson, 2009; Logan, Holcombe, Ryan, Manlove, & Moore, 2007), but
50 unintended pregnancies and abortions have been linked with sexual repeat victimization (Wyatt et al. 1992). Consequently, such a variable may also function as a childhood development risk factor for repeat sexual victimization Post -Sexual Victimization Trauma With respect to childhood development factors that may relate to post -sexual victimization mental health, childhood physical abuse potentially establishes trauma that predates experiences of sexual victimization (Briere & Jordan, 2009). Another pers onal history variable is p hysical health problems which are more likely among survivors of sexual assault or rape than non -victims (Cloutier, Martin, & Poole, 2002) Context of Victimization Sexual Repeat Victimization Perpetrators connect the childhood development and context of the first victimization variables reacting to perceived victim vulnerability and probability of sanctions from formal and informal systems of justice. Factors related to the likelihood of perpetrators acting aggressively and avoiding sanctions include immediate antecedents to the victimization, such as alcohol use or fighting, or direct consequences of the sexual re peat victimization, such as an injury or pregnancy. Although there are some studies contradicting the relationship between substance use and risk of sexual repeat victimization likelihood, overall the literature reveals a fairly strong association (Breitenbecher, 2001; Krebs, Lindquist, Warner, Fisher, & Martin, 2009). Pregnancy has been documented both as a risk and pr otective factor for sexual repeat victimization, so it is unclear how the impact of a pregnancy as a result of a primary victimization influences
51 for sexual repeat victimization (Collins, 1998; Wyatt, 1992) The use of physical violence as a coercion tacti c in a primary victimization predicts future repeat sexual victimization (Casey & Nurius, 2005; Griffin & Read, 2012). An important covariate of coercion tactic is victim offender relationship, with research documenting intimate offenders using greater physical violence than strangers and acquaintances (Murphy et al. 2011). Post -Sexual Victimization Trauma V ariables from the context of victimization may also relate to long -term mental health trauma. A younger age at the time of a primary sexual victimiz ation has been associated with greater long term psychological distress (Masho & Ahmed, 2007; McCutcheon et al., 2010). A minimal number of studies have examined the relationship between victim and /or perpetrator substance use and psychological outcome, but the studies that have examined these factors have found non -significant effects (Campbell et al., 1999; Resnick, Yehuda, & Acierno, 1997). This may be due to the lesser degree of violence used when substances are used as a coercion tactic. H owever, if an individual willingly used alcohol or drugs prior to the assault, they may, due to victim blaming feel a higher degree of guilt or responsibility for their victimization, compounding depression, and/or other negative mental health outcomes ( Brown et al., 2009; Tucker et al., 2005). The nature of the victim -offender relationship has been found in some research to not directly relate to mental health outcome (Campbell et al., 1999; Ullman & Filipas, 2001b ), although other studies have noted that sexual assault by acquaintances or known offenders had a higher association of PTSD (Masho & Ahmed, 2007; Temple et al., 2007). A resultant pregnancy from the first victimization is an additional example of context of the first victimization factor that m ay increase post repeat victimization
52 trauma, due to the both short -term and/or long -term necessity of addressing unwanted pregnancy. Another individual level variable includes maladaptive strategies of coping behavior, such as socially withdrawing which can include staying at home or disengaging from the outside world, which are associated with higher levels of depression post -victimization (Frazier et al. 2005; Gutner et al. 2006; Ullman, et al., 2007). Social P ower Sexual Repeat Victimization Socia l power variables would include those that pertain to one's social and financial resources that could act as protective or risk factors in one's likelihood of being sexually revictim ized or developing severe trauma. Low socioeconomic status and education levels have been documented as a risk factor for sexual and interpersonal victimization and repeat victimization ( Acierno Resnick, & Kilpatrick, 1997; Mears Carlson, Holden, & Harris 2001). Race and ethnicity has been minimally researched with respect to repeat victimization (Classen et al., 2005) One exception includes a study by Urquiza and Goodlin Jones ( 1994) who documented the majority of African American women (62%), just under half of White women (44%) and Hi spanic/Latina women (40%), and one quarter of Asian American women reporting sexual revictimization. Coping strategies and access to social support have primarily been researched in relation to post -sexual victimization recovery; however both variables may be linked indirectly to sexual repeat victimization through the variable of post -victimization distress. Whereas situations where a victim file s a police report, presse s charges, and the perpetrator was
53 convicted have been documented as protective variabl es in partner violence repeat victimization (Jordan, 2004; Lewis, 2004), such criminal justice outcome variables have not been examined to the same extent with respect to repeat sexual victimization. Sexual Victimization Trauma Race /ethnicity has also been found to have some influence (Burge, 1988; McFarlane et al., 2005) or non -significant findings (Elliott et al. 2004) on mental health outcomes. Economic stress has been linked with psychological trauma as a result of victimization; although the relationship between education level and psychological distress post -victimization is less clear (Campbell et al., 2009). Acce ss to formal sources of support, or criminal justice and medical care professionals, has been associated with greater PTSD sy mptomology, potentially reflecting a more traumatic assault, spurring survivors to reach out to multiple forms of social support (Masho & Ahmed, 2007). Generally, negative social reactions are more common for women talking with formal sources of support than informal (Ullman & Filipas, 2001a). However, such variables may also act as protective or risk factors for mental health problems trauma, depending on victim satisfaction with the therapeutic or criminal justice process. Implications of Ecological Model s of Sexual Repeat Victimization and Post -Sexual Victimization Trauma Compared with predominant models of repeat victimization, the ecological model considers not only individual victim and situational factors, but the larger social context influencing how society and offenders react to victim psychology interpret victim behavior, and foster or constrain the social settings conducive to sexual victimization
54 The ecological model extends beyond the context of the victim and the abuse, but additionally consi ders each following disclosure where the victim is informed of who is responsible for her victimization, and whether that victimization was even a crime. Moreover, the ecological model suggests policy intervention at multiple levels, where rape defense cla sses and dry campuses suggested in individual level models are complemented with comprehensive sexual education addressing violent and healthy intimacy in K 12 settings, expansion of sexual assault nurse (SANE) programs and rape crisis programs, and large -scale public health campaign s addressing behavioral and sociocultural norms of sexual victimization and victim blame. Summary Predominant models of repeat victimization focus on individual victim psychology/trauma, victim behavior, or situational factors with little empirical support and policy implications that emphasize the victim. Likewise, PTSD models of post sexual victimization mental health primarily focus on victimization itself as a source of trauma Alternately, ecological models of repeat sexua l victimization and post victimization trauma consider not only victim and victimization specific variables, but additionally the societal reaction to victimization, cultural assignment of responsibility, and access to post victimization support and justic e. The present study proposes an ecological framework for understanding sexual repeat victimization and post victimization mental health, which encompasses individual, situational, and sociocultural conceptual groupings of variables Whereas this framework does not neatly predict how each individual variable impacts the outcome variable, it does organize multiple factors into conceptual groupings The ecological structure is suitable for a complex, ongoing
55 social phenomenon such as sexual repeat victimization and psychological trauma Ultimately any theory of sexual repeat victimization that solely identifies variables distinguishing those victimized by one perpetrator from those victimized multiple times by separate perpetrators will be insufficient, as a c omplete theory must explain the causal mechanisms behind each variable s cumulative contribution to increasing likelihood of repeat victimization (Marx et al., 2005). However, the current research contributes to the first step of theory development: identi fying multiple risk and protective factors of sexual repeat victimization. The present study is not a test of the ecological model due to the limits of the data set employed and instead examines the influence that ecological individual, environmental, and socio cultural variables have on sexual repeat victimization likelihood and post -victimization mental health
56 CHAPTER V RESEARCH METHODS Research Questions In order to glean a more accurate representation of the dynamics behind sexual repeat victimization and post -victimization mental health the present study incorporates a large random -digit dialing national sample with the objective of greater external and statistical validity. The following section begins with a discussion of the study design, followed by the operationalization of the dependent, independent, and control variables. This study focuses on the following two research questions: 1) Which ecologically based factors predict respondents victimized by one perpetrator only versus repeat victims of sexual violence? 2) Which ecologically based factors predict depression symptoms in respondents victimized by one perpetrator only versus repeat victims of sexual violence? Study Design The present research assessed data collected in Patricia Tjaden and Nancy Thoennes research titled Violence and Threats of Violence Against Women and Men in the United States 19941996, known as the National Violence Against Women Survey (NVAWS).1 Data were collected through random -digit dialing (RDD) of households with a functioning phone line in the United States, stratified by U.S. Census region. Women and men 18 years of age or older and residen ts of the households were eligible to participate in the interview with the female version of the survey implemented between 1995 and 1996, and the male version in 1996. Both Spanish and English versions of the 1 Institutional review board (IRB) approval was granted in May of 2011 to conduct secondary data analysis on these data.
57 interview were avai lable and Spanish versions of both male and female surveys were fielded in 1996. D ue to the nature of the survey, female participants were interviewed by female interviewers only. In addition, interviewers were trained on detecting signs that the responden t was apprehensive about being overheard or was in distress, and contact information for local victim support services was offered to respondents disclosing violent victimization. Schulman, Ronca, Bucuvalas, Inc. (SRBI), a professional survey research organization, conducted the interviews using a computer assisted interviewing system. Although the NVAWS includes interviews with males, the current study is restricted to analysis of data from 8,000 female participants due to low rate s of male disclosures of sexual repeat victimization. In addition, women's risk of sexual repeat victimization is almost twice that of men (Barnes Noll, Putnam, & Trickett, 2009) indicating gendered trajectories of sexual repeat victimization Of the women contacted who were eligible to participate in this study, 72.1% agreed to be interviewed. Respondents were asked about demographic information, healthrelated variables, the number of current and past intimate cohabitating relationships and experiences with violent victim ization Sections of the survey covered 1) fear of violence and fear management 2) emotional abuse experienced by intimate partners, 3) forcible rape or stalking, and 4) incidents of threatened violence experienced by any type of perpetrator. If a respond ent indicated during the interview screening that they had experienced attempted or completed rape, investigators followed up with questions on the characteristics and consequences of the most recent victimization incident by perpetrator type. Perpetrator type was identified by the nature of the relationship between the perpetrator and the respondent: current spouse, ex-spouse, male live in partner, female
58 live in partner, other relative, acquaintance, or stranger. Questions regarding detailed information o f each victimization by perpetrator type included the number of times the victimization occurred, if the incident occurred in the last 12 months or in years prior to the interview, whether drugs or alcohol were used, whether injuries or pregnancy occurred, and criminal justice system engagement and outcome. Sample The present study contributes to the literature through its use of a national, random -digit -dialing sample of women 18 years of age or older. Desai et al. (2002) note that majority of research linking childhood and adult victimization has been carried out with clinical college, and convenience samples In contrast, population -based studies are able to provide an estimation of victimization incidence and consequently establish a basis for genera lizable intervention programs. Women in college and clinical samples share similar education levels and/or socioeconomic statuses, which may influence the probability of repeat victimization and post victimization mental health (Messman -Moore & Long, 2003 ). As prior research has indicated, low socioeconomic status (Ellis Atkeson & Calhoun, 1982; Mears et al., 2001; Miller Moeller, Kaufman, DiVasto, Pathak, & Christy, 1978), race (Mears et al., 2001), and transiency (Ellis et al., 1982; Miller et al., 197 8) are risk factors for repeat sexual victimization and post -victimization mental health, which are potentially truncated or overrepresented in samples e xclusively comprising college students or clinical populations respectively Additionally, the complex ity of the questionnaire used in the NVAWS includes a range of variable constructs permitting the selection of multiple ecologically based variables Lastly, the
59 NVAWS is one of few publicly available data sets with a sufficient number of sexual repeat vi ctims to conduct multivariate analysis. Despite myriad advantages of the NVAWS data, responses were collected over 10 years ago. As a result, it is important to consider whether the relationships among variables collected in the mid 1990s would be applica ble to the present day. Variables were selected based on ecological assumptions of power, with the understanding that gender and race/ethnicity influence inequality. In addition, it is important to consider whether the incidence and scope of sexual victimi zation has changed significantly over the past 20 years As a result, the sample description will include a comparison to present -day national demographics. Wi thin the total sample of women the average age was 44.2 (SD = 17.57) (See Table 5.1) The h ighest education al attainment levels for respondents were most commonly high school (34.6%), or attending some college (29.3%). Consistent with the sample, presently 29.3% of women ages 18 and older in the civilian non -institutionalized population gradua t e high school and 29.8% attend some college (U.S. Census Bureau, 2012). Over half of women earned an income of $20,000 or less (57.3%). When considering the generalizability of the NVAWS sample to todays context, it must be noted that the female to -male earnings ratio has increased from 70% in 1995 to 77% in 2008 (DeNavas -Walt, Proctor, & Smith, 2009). Progress in the gender earnings inequality gap may arguably result in diminished generalizability of the present research. However, t he Gini index grew from 40.8 in 1997 to 45 in 2007 (C entral Intelligence Agency, 2013), indicating on a scale of 0 to 100, that inequality in family income has increased between the point in time when the NVAWS was collected and the present day. Shaefer and Edin ( in press ) document this economic trend noting that in
60 1996, 1.7% of households lived in extreme poverty, or on less than $2 per family member per day, in contrast to 4.3% in 2011. Table 5. 1 : Sample Characteristics. Demographic Variabl es Frequency Percentage Age 18 25 947 12.1% 26 35 1773 22.6% 36 45 1885 24.0% 46 55 1396 17.8% 56 65 840 10.7% 66 and Older 1014 12.9% Marital Status Single 1224 15.4% Married/Common Law 4999 62.9% Divorced/Separated/Widowed 1730 21.8% Race White and Non Latina 6217 79.2% White and Hispanic or Latina 235 3.0% Black or African American 780 9.9% Asian or Pacific Islander 133 1.7% Native American or Alaskan Native 88 1.1% Mixed Race 397 5.1% Education Level Some Grade School 856 10.7% High School Graduate 2752 34.6% Some College 2336 29.3% College Graduate 1360 17.1% Post Graduate 659 8.3% Employment Full Time 3678 47.6% Part Time 1009 13.1% Homemaker 1233 15.9% Unemployed 278 3.6% Student 357 4.6% Retired 1176 15.2% Individual Income Less than $20,000 3395 57.3%
61 Demographic Variabl es Frequency Percentage $20,000 $35,000 1423 24.0% $35,000 $50,000 719 12.1% $50,0000 $80,000 329 5.5% Greater than $80,000 62 1.0% Rape Victimization History Never Victimized 6592 82.5% Victimized by One Offender 1241 15.5% Victimized by Multiple Offenders 159 2.0% Most women in the sample identified themselves as white, nonHispanic (79.2%) or Black or African American ( 9.9%) Racial and ethnic demographics have markedly changed since the 1990s T he proportion of W hite Americans in 1990 compris ed 77% of the population, and only 72.4% in 2010, and 9% of Americans in 1990 were Hispanic or Latino versus 16.3% in 2010 (U.S. Census, 1992a 1992b, 2011 ). However, r acial and ethnic inequality has not progressed substantially over the past 15 years. In 1995, 11.2% of White, non Hispanics, 29.3% of African Americans, 30.3% of Latinos, and 14.6% of Asian -Americans lived below the poverty level, consistent with 2009, whe re 12.3% of White, non Hispanics, 25.8% of African Americans, 26.6% of Latinos, and 12.1% of Asian Americans lived in poverty (U.S. Census, 2012). One thousand four hundred and eight women indicated in the NVAWS rape screening question that they had experi enced a completed or attempted rape, and of the women who went on to answer questions about each victimization (1 400), 15.5% were raped by one offender and 2% were raped by more than one offender over their lifetime. The Bureau of Justice Statistics indicates that between 1995 and 2010, attempted or complete d rape or sexual assault against women and girls 12 and older decreased from 5.0 per 1,000 women and girls to 2.1 (Planty, Langton, Krebs, Berzofsky, & Smiley -
62 McDonald, 2013). Rates of sexual viole nce against men also decreased from 0.6 to 0.1 over that same time period, indicating a closing gender gap in sexual violence, but a gender gap nonetheless. Conversely, a lower percentage of sexual crimes reported to the police culminated in arrests in 2005 10 (31%) in comparison to 199498 (47%) and the rates of reporting to police increased minimally from 29% between 1994 1998 to 35% in 2010 (Planty et al., 2013) When considering changes in the scope of sexual victimization over time, Planty et al.'s (2013) recent assessment indicates few trends. Sexual victimizations committed by strangers were consistently in the minority at around 22% between 1994 and 2010. In addition, over time, women and girls who were never married, divorced, or separated and wom en in the lowest socioeconomic status consistently experienced the highest rates of sexual victimization, compared with married or widowed females, or women and girls in higher income brackets. The location of the crime was most consistently close to or at the victims home, and primarily occurred when the victim was sleeping or doing other activities in the home across time periods (Planty et al., 2013) One trend over time specifies that in 19941998 women and girls in urban areas were at greatest risk, compared with those in suburban or rural areas. In 20052010, women and girls in rural areas become the most vulnerable (3.0 per 1,000), followed by urban (2.2 per 1,000), and then sub urban areas (1 8 per 1,000) (Planty et al. 2013). This shift in geographic vulnerability may reflect the increasing concentration of poverty in rural America over time (Farrigan, 2012). Although improvements in gender income equality may have influenced w omens social power, increased overall inequality would suggest that an ecological model is appropriate where gender alone is not representative of less power,
63 and instead, an intersection of social identities such as gender, race, and socioeconomic status may leave particular subgroups of individuals vulnerable to crime and excluded from systems of justice. Measures Dependent Variables Research Question One: Predictors of Rape Victimization by One Perpetrator Versus Multiple Victimizations by Separate Perpetrators The measure for the dependent variable of sexual victimization included a series of questions about an experience where someone used or threatened force to penetrate the respondents vagina or anus by penis tongue, fingers, or object, or the respondent s mouth by penis. Both attempted and completed rape incidents were included. Although the bulk of research on repeat sexual victimization assesses sexual assault, which includes unwanted sexual touching or exposure in addition to attempted and completed rape, the current study examines attempted and completed rape only, due to the particular measure used in the NVAWS. However t he isolation of rape as a dependent variable may enhance the current researchs abi lity to examine the phenomenon of repeat victimization as there may be a more consistent likelihood of repeat victimization when the initial assault is more severe (Mayall & Gold, 1995; Roodman & Clum, 2001), and risk factors for rape versus other types of sexual victimization may vary (Testa & Dermen, 1999). Respondents were asked behaviorally specific questions about sexually violent acts in order to obtain greater accuracy in responses, as some research indicates that when
64 individuals are asked if they have been raped or sexually assaulted, lower rates of victimization are reported despite having experienced the crimes (Fisher & Cullen, 2000). Those who had experienced one or ongoing rape victimization by one perpetrator over a lifetime were categor ized as one perpetrator rape victims and coded as (0), and those who had experienced two or more rape victimizations by distinct perpetrators over a lifetime were categorized as repeat victims and coded as (1). The operationalization of repeat victimization as multiple victimizations by different offenders was selected for methodological reasons. In the survey, details regarding the most recent victimization by each distinct offender were collected, hindering the ability to isolate characteristics of the first victimization by offenders who assaulted the victim multiple times. As a result, in order to preserve time order, repeat victimization was operationalized as multiple victimizations by different offenders However, survey research generally has difficulty isolating unique characteristics of each incident when multiple victimizations were perpetrated by the same offender. As a result, the majority of sexual repeat victimization research operationalizes repea t victimization as two or more sexual victimizations perpetrated by distinct offenders (Casey & Nurius, 2005). More specifically, the dependent variable comprises two groups. T he first group, one -perpetrator rape victims includes women who have experienced one or ongoing rape victimizations by one perpetrator only. The second group includes repeat victims, consisting of women sexually victimized by distinct perpetrators on separate occasions (See Figure 5.1) Including women who have experienced on e or ongoing rape victimization by one perpetrator in the non repeat victimized group appears counter intuitive. However, the differentiation between non repeat victim ized respondents and
65 repeat victims is not based exclusively on the number of incidents Rather it concerns the number of perpetrators, consistent with previous research and operationalizations of sexual repeat victimization (Casey & Nurius, 2005). Only rarely has ongoing victimization by one perpetrato r been isolated from one violent incident by one perpetrator in repeat victimization measurements (Casey & Nurius, 2005). Number of Perpetrators Number of Rape Victimizations One Perpetrator Victims 1 1 or ongoing Repeat Victims More than 1 More than 1 Figure 5.1: Operationalization of Sexual Repeat Victimization A total of 1,408 female sexual assault victims were identified. Of the 1,408 women who reported an attempted or completed rape in the screening questions, 1,400 respondents gave additional details on their victimization experiences comprising 1,241 women raped by one perpetrator and 159 women raped multiple times by separate perpetrators Due to the fact that time since victimization was measured in years, 19 repeat victims who reported vic timizations in the same year were excluded due to the inability to determine whi ch victimization was the first. In total then, the examination of predictors of one -perpetrator versus repeat rape victimization was based on a sample of 1,381 females age 18 o r older. Research Question Two: Long T erm Post -V ictimization Mental Health An eight question inventory examine d longterm post -sexual victimization depression. The time frame encompassed the week prior to the interview, inquiring how often the respondent felt a ) full of pep, b ) very nervous, c ) down in the dumps, d ) full of energy, e ) downhearted and blue, f ) worn out, g ) happy, and h ) tired. The response ranges
66 for the depression inventory were coded ordinally as: never (0), rare ly (1), some of the time (2), and most of the time (3). Questions a ), d ), and g ) were reverse coded so that answers indicating more pep, energy, or happiness were given lower values on the depression scale. A greater number of questions answered with a sco re of three or four indicated more frequent and extensive depression symptoms Answers to the eight questions were summed and averaged for a total possible continuous range of 0 1, with 1 indicating the highest level of depression, a nd 0 indicating no depr ession symptoms in the week prior to the interview 80 for the depression scale, indicating strong internal reliability. Independent Variables Independent variables in this study were chosen based on prior repeat sexual victimization and post -victimization distress research, and organized according to the ecological framework. Variable selection was guided by the assumption that reduced individual control, through financial, institutional, prejudicial, physi cal, or psychological constraints, heightens the risk of sexual repeat victimization and post -victimization distress. This assumption reflects intersectional theories of sexual victimization (Bograd, 1999; Crenshaw, 1991; Heise, 1998). Independent variable s vary for r esearch q uestion o ne and r esearch q uestion t wo in the current research due to temporal issues. Research question one considers predictors of sexual repeat victimization, and as a result, independent variables must reflect the time period prior to the most recent victimization, or represent ascribed characteristics, such as race/ethnicity or socioeconomic status. Research question two examines mental health at the time of interview, and consequently a b roader selection of independent variables may be considered.
67 Childhood D evelopment Childhood development variables include factors from one's individual history that are predicted to influence likelihood of repeat victimization and long -term post victimization mental health (See Tables 5. 2 and 5. 3 ). For research question one and two early pregnancies or giving birth prior to the age of 18, and childhood maltreatment are childhood development independent variables. In addition, for research question two on post victimization mental health rape victimization history is an independent variable. It should be noted that childhood development variables can act as risk or protective factors differentiating not only repeat rape victims from those victimized by one perpetrator only but also non -victims from rape victims. Although research has associated childhood maltreatment and early parenthood with sexual victimization (Boyer & Fine, 1999; Sanders & Moore, 1999) and repeat sexual victimization (Schaaf & McCane, 1998; Wyatt et al. 1992), littl e work has been done on the relative strength of the relationships, so it is difficult to speculate whether these variables have an equivalent or varied influence differentiating non -victims from rape victims versus repeat victims from those victimized by one perpetrator only One exception includes Merrill et al. (1999), who found that once childhood sexual assault was controlled for, childhood physical abuse did not independently influence likelihood of adult sexual assault, potentially indicating childhood maltreatment is a weaker predictor of repeat sexual assaults than a previous sexual assault.
68 Table 5. 2 : Ecological Variables in Sexual Repeat Victimization Likelihood Model Variable Type Variable Control Variables Years since rape victimization by first perpetrator Age at rape victimization by first perpetrator Childhood development Parenthood at < 18 years of age No children Childhood maltreatment First perpetrator victimization context Victim offender relationship (reference: Family) Stranger Acquaintance Current or Ex Partners Hanging out prior to victimization Drinking/drugs prior to victimization Date prior to victimization Fighting prior to victimization Offender use of verbal or physical coercion Whether victim got pregnant from rape Social power Race/ethnicity Income Education Disclosure of victimization by first perpetrator to therapist Disclosure of victimization by first perpetrator to informal sources of support Degree of police involvement regarding victimization by first perpetrator Table 5 3 : Ecological Variables in Post -victimization Mental Health Model Variable Type Variable Control Variables Years since rape victimization by first perpetrator Childhood development Rape by one offender versus multiple rape victimizations by separate offenders Chronic disease or serious injury First perpetrator victimization context Age at rape victimization by first perpetrator Victim offender relationship (reference: Family) Stranger Acquaintance Current or Ex Partners Hanging out prior to victimization Drinking/drugs prior to victimization Date prior to victimization
69 Variable Type Variable Fighting prior to victimization Offender use of verbal or physical coercion Whether victim got pregnant from rape Whether victim needed time off from work or school post assault Social power Race/ethnicity Income Education Disclosure of victimization by first perpetrator to therapist Disclosure of victimization by first perpetrator to informal sources of support Degree of police involvement regarding victimization by first perpetrator Early parenthood is determined by subtracting the age of each respondent s oldest child from the age of the respondent at the time of interview. This is a categorical variable, including women who never had children (0 ), women who gave birth prior to 18 years of age (1) and women who became mothers at age 18 or over (2 ). For the regression, giving birth at 18 and over is the referent group, with giving birth prior to 18 years of age and never having children coded as dummy variables. Childhood maltreatment is measured with a twelve item scale, adopted from Strauss Conflict Tactics Scale (Straus, Hamby, Boney-McCoy, & Sugarman, 1996) where respondents were asked whether a parent, step parent, or guardian ever a ) threw something at them, b ) pushed, grabbed, or shoved them, c ) pulled their hair, d) slapped or hit them, e ) kicked or bit them, f ) choked or attempted to drown them, g ) hit them with some object, h ) beat them up, i ) threatened them with a gun, j ) threatened them with a knife or other weapon, k ) used a gun on them, or l ) used a knife or other weapon on them beside a gun. For each item, respondents were coded as 1 if they had experienced that behavior and a 0 if not. Responses for the
70 twelve childhood ma ltreatment measure we re summed and averaged resulting in a continuous measure that ranges from 0 1. Higher values indicate more experiences with childhood maltreatment. Chronic and severe physical health problems were determined by two questions : "Have y ou ever sustained a serious injury, such as a spinal cord, neck or head injury that is disabling or interferes with your normal activities?" and "Do you have a chronic disease or health condition that is disabling or interferes with your normal activities? Both questions had "yes" (1) or "no" (0) response options. The questions were collapsed into one measure, where an answer of "yes" (1) to one or both questions reflect ed a history of physical health problems, and an answer of "no" (0) to both questions indicated no history of physical health problems. Context of Rape Victimization Variables Independent variables measuring the context of the most recent rape victimization by the first or only offender include victim age, the pre -victimization context, coercion tactic used during the assault, the perpetrator -victim relationship, and pregnancy as a result of the rape for both the repeat victimi zation and mental health models (See Tables 5. 2 and 5. 3 ). Regarding the pre victimization context, respondents were asked "Could you tell me what started or triggered this incident?" Answers included yes (1) or no (0) responses to four options: a) Nothing in particular or hanging out, b ) Perpetrator and/or offender using substances c) On a date or engaging in sexual activity, or d) Fighting or breaking up. Coercion tactic wa s determined by whether the perpetrator a ) slapped, b ) hit, c ) kicked, d ) bit, e ) choked, f ) attempted to drown, g ) beat up, or h ) hit the respondent with an object, and h ) whether the perpetrator used a gun, knife, or other weapon. Threat
71 occurred if the perpetrator a ) threatened to harm or kill the respondent or someone close to them, or b ) if the respondents believed someone close to them or themselves would be seriously harmed or killed during the assault. Respondents answered yes (1) or no (0) to each violence and threat question. Violence and threat of violence were collapsed into one ordinal measure indicating w hether the offender used verbal coercion (0), threats of violence (1), or physical violence or a weapon (2). Pregnancy as the result of the first sexual assault was measured by a yes (1) or no (0) answer. For the post -victimization mental health model, respondent age at the time of the rape victimization by the first or only offender is measured in years. Perpetrator -victim relationship was identified by the respondent selecting whether the perpetrator was a) a current spouse, b) an ex -spouse, c) a male live -in partner, d) a female live in partner, e) a relative, f) an acquaintance, or g) a stranger. Responses were collapsed into four nominal categories including strangers (0), acquaintances, (1) current or ex partners, (2 ), or family members ( 3 ). For th e regression, family wa s the referent group, with strangers, acquaintances, current or ex -partners, and family members coded as dummy variables. Social P ower Social power variables include race/ethnicity, income, education, disclosure to a therapist, disclosure to an informal source of support, and degree of engagement with the criminal justice system for research questions one and two (See Tables 5. 2 and 5. 3 ). Race and ethnicity we r e determined through yes (1) or no (0) responses to six response options: a) White/Non Hispanic, b) Black or African -American, c) Asian or Pacific Islander, d) American Indian or Alaskan native, e) Mixed race, and f) Hispanic. Due to
72 88% of respondents identifying as White/non-Latina and African -American, the variable wa s co llapsed into a dichotomous measure, with Latina and women of color represented by 0, and non-Latina White women represented by 1. The income measure inquired how many earnings the respondent accrued from all sources in 1995 before taxes. Although househol d rather than individual income is traditionally measured in economic research, the present study is primarily interested in the respondent herself, rather than household buying capability. Although the income measure reflects income at the time of intervi ew, the Pew Charitable Trust finds that 70% of Americans born in the bottom quintile of family income remain below the middle as adults, reflecting a degree of economic stability over a lifetime (Urahn, Currier, Elliott, Wechsler, Wilson, & Colbert, 2012). Considering individual income is additionally advantageous when examining interpersonal violence, as the wage gap between men and women has been associated with higher rates of domestic violence (Aizer, 2010). Responses include 10 options: less than $5,000 (1), $5,000$10,000 (2), $10,000$15,000 (3), $15,000$20,000 (4), $20,000$25,000 (5), $25,000$35,000 (6), $35,000$50,000 (7), $50,000$80,000 (8), $80,000$100,000 (9), or over $100,000 (10). Although these response options are not mutually exclusive the options were read to respondents over the phone and r espondents were instructed to stop the interviewer once they heard the category that applied. As a result, one could assume respondents interpreted the categories to be: less than $5,000 (1), $5,000 $10,000 (2), $10,001$15,000 (3), $15,001$20,000 (4), $20,001$25,000 (5), $25,001$35,000 (6), $35,001$50,000 (7), $50,001$80,000 (8), $80,001$100,000 (9), or over $100,001 (10). Education level determined the highest level that the respondent compl eted, including seven categories ranging from
73 no schooling (1), first through eighth grade (2), some high school (3), high school graduate (4), some college (5), four year college degree (6), to postgraduate (7). Similar to income, responses were coded so that higher values reflected higher education levels. A disclosure to a therapist about the rape victimization by the first or only offender is coded as yes (1) or no (0). Access to i nformal source s of support was indicated by whether, in response to the question "To whom did you talk to about this incident?", the respondent answered a) family, inlaws, b) friend/neighbor, c) minister/clergy/priest/rabbi, d) intimate partner, or e) a coworker, boss, employer. Respondents gave up to four answers to thi s question, and as a result, informal support was coded as an ordinal variable between one and four from told no one (0) to told four sources of informal support (4). Victim engagement of the criminal justice system was an ordinal variable with three r esponse options, no police report (0), police report (1) and charges filed (2). Control Variables Control variables include d years since the most recent or only rape victimization by the first or only offender or the time between the interview and the victimization, ranging from 0 years to 97 years and age at primary rape victimization Recency of sexual victimization is a commonly included control variable in sexual victimization research, as the length of time since a trauma influences memory info rmation access (Higgins, 1989). Age at rape victimization wa s incorporated as a control variable in conjunction with the recency variable in the repeat rape victimization model analysis to account for variability in exposure time over a lifetime in which an individual may be sexually victimized.
74 Data Analysis Prior to examining multivariate models, data we re first assessed descriptively and relationships were considered at the bivariate level. Subsequently, the data we re analyzed at the multivariate level to answer the first research question: 1) Which ecologically based factors predict respondents victimized by one perpetrator only versus repeat victims2 of sexual violence? A logit regression model wa s applied to researc h question one. The Hosmer and Lemeshow R2-statistic was examined to determine the amount of variance in the dependent variable explained by the independent variables, and the sexual repeat victimization likelihood models level of explanatory power The s ignificance of each coefficient wa s interpreted, as well as the odds ratios for significant variables. All statistical significance tests we re linear regression indicated the average variance inflation factor (VIF) was 1.25, with the VIF minimum = 1.1 and the VIF maximum = 1.37. These VIF scores determined multicollinearity is not problem atic (Bowerman & OConnell, 1990). A VIF of 1 indicates no correlation between one predictor and other predictors. Generally, VIF scores greater than 4 necessitate additional investigation of the variables relationships, and VIF scores greater than 10 ref lect serious multicollinearity concerns (Bowerman & OConnell, 1990). With respect to the regression rule of thumb recommending a minimum of 10 observations per predictor (Harrell, 1984), and the high number of independent variables, models were run initia lly with all predictors, and then re -specified to only include variables significantly different among groups in the preliminary bivariate 2 Again, repeat victimization does not refer to the number of incidents but to the number of perpetrators, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
75 analyses. The significance of individual coefficients as well as the overall fit of the model were consistent between models, and as a result the full models were considered. Ordinary least squares (OLS) regression wa s used to investigate research question two, whether ecologically based factors in a model of long term post -victimization mental health vary for responden ts victimized by one perpetrator only and repeat victims of sexual violence OLS regression wa s applied to the full sample of victims first and then separately for those victimized by one perpetrator and repeat victims. Significant independent and control variables we re interpreted based on regression coefficients, and Adjusted R2 determined the explanatory power of the ecological model for post -sexual victimization mental health. Missing Data Analysis For variables with missing data, multiple imputation wa s applied. Multiple imputation wa s selected due to the advantages of this technique adding random normal error to each imputed estimate, in contrast to single imputation or the expectation maximization algorithm (Graham, 2012). Specifically, a Bayesian procedure, the MarkovChain Monte Carlo (MCMC) method impute d missing data based on the non monotone, arbitrary patterns of the missing data. The MCMC method of multiple imputation is the most widely used multiple imputation strategy, due to its ability to ha ndle almost any pattern of missing data and its computational efficiency (Allison, 2009). Although complete case analysis or listwise deletion is a traditional approach to missing data, listwise deletion can result in biased parameter estimates and a loss of power Graham (2009) recommends against listwise deletion even in data sets with minimal missing data or missing data that are not missing at random. According to
76 Allison (2009), no test is available to detect whether data are missing at random. Howeve r, the inclusion of auxiliary variables or variables not intended for analysis but "helpful" to the imputation, in the multiple imputation can diminish the level of error that results from imputing data that are not missing at random (Graham, 2012). In th e present study, a total of 62 variables we re imputed, with an average of 1.92% missing data (See Table A 1), and 27 additional auxiliary variables ar e included as predictors in the multiple imputation procedure which had no missing data For each iteration and variable included in the multiple imputation model, a fully conditional specification method estimate d a model with the other variables as predictors in order to impute missing data. After estimating predicted values, random draws we re taken from the estimated error distribution for each linear regression. The estimation of random variation addresses the bias present in most imputation methods, where variance is underestimated (Allison, 2009). Although scholars previously suggested ar ound five imputations were sufficient, more recent research estimates a minimum of twenty to forty imputations to obtain statistical power equivalent to full information maximum likelihood analyses (Graham, Olchowski, & Gilreath, 2007). The present analysi s calculat ed missing values estimates across 200 iterations and 50 imputations. After generating the fifty imputed data se ts, and analyzing the data, results we re combined to obtain multiple imputation inference using Rubins rules (Rubin, 1987). Parameter s we re estimated through averaging the regression coefficients over the number of imputed data sets, or fifty Standard Errors (SE) we r e calculated through first estimating the within imputation variance ( U ), or the average of the squared SE over m, or the number of imputations, 2 / m,
77 as well as the between imputation variance ( B ), or the sample variance of the regression coefficient (P) over the 50 imputed data sets, B = S2 P. Selection Bias In addition to the challenges of incomplete cases, the current data present ed another potential problem: censored observations on the dependent variable. A classic example of censored observations is the assessment of the relationship between income and voting preference, where the absence of non-voters i n the sample misleads conclusions about income and voting behavior in the general population (Dubin & Rivers, 1989). In the present research, a sample comprising women sexually victimized by one or multiple perpetrators excludes non -victims, despite the possibility that factors differentiating non -victimization from rape victimization may relate to those differentiating repeat victimization from victimization by one perpetrator only For example, alcohol consumption has been separately documented as a risk factor for both sexual victimization and sexual repeat victimization (Abbey, 2002; Greene & Navarro, 1998; Kaysen, Neighbors, Martell, Fossos, & Larimer, 2006; Testa et al. 2010). Likewise, some researchers assessing risky sexual behavior observe an assoc iation with both sexual victimization and sexual repeat victimization (Champion, Foley, Durant, Hensberry, Altman, & Wolfson, 2004; Fargo, 2009; Himelein, 1995; Krah et al. 1999). However, a recent examination of predictors of two sexual victimization s tates: non victim or victim and single or repeat victim, indicated that the two states were unrelated (Fisher et al., 2009). Heckman (1974) proposed a method for testing and addressing selection bias in linear regression, and Dubin and Rivers (1989) extended this procedure
78 to logit and probit models. In the current research, a logit regression model predict ed repeat victimization likelihood ( ) based on multiple predictors, such as childhood maltreatment and social support (denoted by ) with some variance based on unknown factors ( ) = + However, a sample of victims of sexual assault may be censored, should the likelihood of repeat victimization be found to be dependent on the likelihood of ever being victimized. As a result Heckman (1974) recommends first modeling a "selection equation, to determine if the likelihood of selection into the population of interest, in this case survivors of rape relates to the likelihood of being selected into the outcome of interest in the substantial equation or repeat rape victimization. A selection equation differentiating likelihood of nonvictimization from victimization ( ) m ay share predictors ( ) in common with those of the substantial equation, such as demographic variabl es, and must include additional factors unique to the selection equation. = + When is greater than zero, victimization has occurred and the number of victimizations ( ) may be observed, whereas a equaling zero indicates n onvictimization and a censored number of victimizations ( ) In this case, a control variable based on the residuals of the selection equa tion, the Inverse Mills Ratio, wa s added as a factor in the substantial equation predicting repeat victimization in order to test and if necessary, correct for the unmeasured influence of any victimization. Lastly, although the Heckman correction procedure provides unbiased parameter estimates,
79 standard errors are biased by heteroscedasticity. As a result, the correc ted standard errors a re calculated through a Weighted Least Squares regression (Smits, 2003). In the present study, two selection models we r e specified for each substantial equation. Each selection model may share predictors with those of the substantial e quation, but must have at least one unique predictor. For the sexual repeat victimization likelihood substantial equation, the selection model into a sample of rape victims includes childhood maltreatment, domestic violence, education, depression, and subs tance misuse (See Table A 4 ). Childhood maltreatment has been associated with sexual victimization (Kimerling et al., 2007), and a history of domestic violence wa s included due to the high risk of sexual victimization in violent relationships ( Frieze & Bro wne, 1989). Likewise, education level wa s incorporated based on the high risk of sexual violence individuals face when attending institutions of higher education (Fisher, Cullen, & Turner, 2000). Measures of depression and substance misuse at the time of interview we re predictors because of their potential relationship with pre -victimization mental health, and subsequent association with victimization. Heritability of major depressi on is estimated to be 34%, with the remaining associations explained by environmental factors (Nes, Czajkowski, Rysamb, rstavik, Tambs, & Reichborn-Kjennerud, 2012). For the sexual repeat victimization likelihood substantial equation the Heckman selecti on equation analyses indicate d that depression (AOR=2.15, se = 0.11, p = 0.001), substance use (AOR=1.68, se = 0.15, p = 0.001), education level (AOR=1.08, se = 0.03, p = 0.01), experiencing childhood maltreatment (AOR=28.03, se = 0.20, p = 0.001) and dome stic violence (AOR=1.30, se = 0.07, p = 0.001) all significantly increased the odds of becoming a victim of sexual victimization rather than a non -victim
80 For the post -victimization mental health substantial equation, the selection model into a sample of victims include d childhood maltreatment, domestic violence, and education (See Table B 9). Initial analyses conducted as a part of the Heckman selection equation indicate d that education (AOR=1.08, se = 0.03, p = 0.001), experiencing childhood maltreatment (AOR= 35.16, se = 0.20, p = 0.001) and domestic violence (AOR= 3.67, se = 0.07, p = 0.001) all significantly increased the odds of experiencing any sexual victimization versus being a non -victim Each selection equation wa s used to formulate the Inverse Mills Ratio, or lambda, for each substantial equation. A significant coefficient for any of the lambdas in the regression equations indicate a selection effect, or that the unobservable variables contributing to selection into a sample of rape victims we re significantly correlated with selection into a sample of repeat rape victims. Summary A large national random -digit -dialing sample wa s assessed to answer the following research questions: 1) Which ecologically based factors predict respondents victimized by one perpetrator only versus repeat victims of sexual violence? and 2) Which ecologically based factors predict depression in respondents victimized by one perpetrator only versus repeat victims of sexual violence? Relatio nships among ecological variables, repeat victimization and depression wer e first descriptively assessed on a bivariate level. Logistic regression addresse d which factors alternately predict rape by one offender versus multiple rape victimizations by sepa rate offenders. Predictors of longterm post rape depression we re considered first among the full sample of victims, and then separately for women victimized by one offender versus women repeat victimized by multiple offenders using OLS regression.
82 CHAPTER VI RESULTS Descriptive and bivariate statistics of the sample are examined first Bivariate and multivariate analyses are then discussed for each research question. All reported statistics describe the data post imputation analysis. Sample Descrip tion The sample is described through comparing respondents victimized by one perpetrator only and repeat victims of rape by separate perpetrators on demographic s and victimization context variables (see Table 6.1 ). Table 6.1: Descriptive Statistics for Sample and Bivariate Comparisons between One -Perpetrator Victims and Repeat Rape Victims (n=1381) One Perpetrator Victims Repeat Victims (n=1241) (n=140) Mean (SD)/ n(%) Mean (SD)/ n(%) Control v ariables Years since victimization by first perpetrator *** 20.99(13.48) 26.24(13.34) Age at rape victimization by first perpetrator *** 18.39(9.11) 14.20(8.67) Childhood development Parenthood No children 311(25.1%) 28(20%) Parenthood at < 18 years of age 107(8.6%) 16(11.4%) years of age 823(66.3%) 96(68.6%) Childhood maltreatment*** 0.16(0.20) 0.26(0.24) Depression*** 0.38(0.19) 0.44(0.21) Domestic violence history*** 0.53(0.50) 0.88(0.33) First perpetrator victimization context Victim offender relationship*** Stranger 188(15.1%) 10(7.1%) Acquaintance 572(46.1%) 37(26.4%) Partner 249(20.1%) 32(22.9%) Family 232(18.7%) 61(43.6%) Hanging out prior to 0.11(0.42) 0.23(0.31)
83 One Perpetrator Victims Repeat Victims (n=1241) (n=140) Mean (SD)/ n(%) Mean (SD)/ n(%) victimization*** Drinking/drugs prior to victimization 0.22(0.40) 0.20(0.38) Date prior to victimization 0.16(0.36) 0.11(0.31) Fighting prior to victimization 0.19(0.38) 0.16(0.36) Offender use of verbal or physical coercion 0.95(0.91) 1.15(0.86) Whether victim got pregnant from the assault 0.05(0.22) 0.01(0.08) Social power Education 4.81(1.10) 4.80(1.13) Race/ethnicity 0.78(0.42) 0.78(0.42) Income 0.91(1.25) 0.75(1.04) Disclosure of victimization by first perpetrator to therapist + 0.30(0.46) 0.38(0.49) Disclosure of victimization by first perpetrator to informal sources of support 0.79(0.73) 0.78(0.68) Degree of police involvement regarding victimization by first perpetrator 0.29(0.71) 0.25(0.66) Notes: + ** Demographic Variables Comparison between women victimized by one perpetrator and repeat victims3 reveal multiple demographic differences Childhood maltreatment varied among the victimization groups F (1,1379) = 28.91, p = .001. Due to significant differences in population variances, the GamesHowell post -hoc analysis procedure was selected, finding respondents victimized by one perpetrator (M=0.16, SD=0.20) were significantly less likely than repeat victims (M=0.26, SD=0.24) to report childhood maltreatment. 3 Again, repeat vict imization does not refer to the number of rape incidents but to the number of perpetrators, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
84 Depression scale scores also differed F (1,1379) = 12.57, p = .0 0 1, such that respondents sexually victimized by one perpetrator (M= 0.38, SD=0. 19) scored lower values than women victimized multiple times by separate offenders (M= 0.44, SD=0. 21) on a depression measure. Giving birth be fore the age of 18 or never having children did not significantly differ between those sexually victimized by one perpetrator and those victimized multiple times by different perpetrators 2(2 )=3.03, p = 25. A history of domestic violence also related to sexual victimization, F (1, 1379) = 63.87, p = .00 0 Due to significant differences in population variances, the Games Howell post hoc analysis procedure was selected, finding that that women victimized by one perpetrator (M=0.53, SD=0.50) were less likely than women victimized multiple times by separate perpetrators (M=0.88, SD=0.33) to report experiencing domestic violence. Race/ethnicity, F (1, 1379) = 0.44, p = .52, education, F (1, 1378) = 0.04, p = .85, and income, F (1, 1379) = 2.33, p = .20, were unconnected to victimization experience. Rape Victimization Variables All control variables varied significantly between respondents victimized by one perpetrator and those raped multiple times by distinct perpetrators, including years since rape victimization by the first or only offender and age at rape victimization by the first or only offender Repeat victims4 were sexually victimized earlier in life (M=14.20, SD=8.67) than women victimized by one per petrator only (M=18.31, SD=9.11) F (1, 1378) =25.74, p = .000. Similarly, r epeat victims victimization by the first offender 4 Again, repeat victimization does not refer to the number of rape incidents but to the number of perpetrators, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
85 occurred longer prior to the interview (M=26.24, SD=13.34) than the victimization of one -perpetrator victims (M=20.99, SD=13.48), F (1, 1378) =15.84, p = .000. Within the context of victimization by the first or only offender the offender was significantly more likely to be a family member and less likely to be a stranger or acquaintance among repeat victims than those victimized by one perpetrator only 2(3 )= 53.92, p = .00 0 Precipitating circumstances were more likely to be described as nothing in particular among women victimized multiple times by separate offenders (M=0.23, SD=0.31) than by those raped by one offender (M=0.11, SD=0.42), F (1, 1378) =18.25, p = .000. Reports of drinking or taking drugs, F (1, 1378) =0.96, p = .65, being on a date or engaging in sexual activity, F (1, 1378) =2.15, p = .17, and fighting prior to the rape victimization by the first or only perpetrator F (1, 1378) =1.09, p = .40, did not vary based on repeat victimization history Repeat victims reported the use of force or a weapon during their first victimization (M=1.15, SD=0.86) more frequently than did one perpetrator victims (M=0.95, SD=0.91) F (1 1378) =6.94, p = .0 1 Unexpectedly, repeat victims were also more likely to become pregnant after the rape perpetrated by the first offender (M=0.05, SD=0.22) than women raped by one perpetrator only (M=0.01, SD=0.08), F (1, 1378) =5.62, p = .0 2 The dif ference in disclosure s to therapists regarding the rape victimization by the first or only perpetrator between one perpetrator victims (M=0.30, SD=0.46) and repeat victims (M=0.38, SD=0.49) broached significance at p = .0 6 F (1, 1378) =3.56, to the effect that women victimized multiple times by separate perpetrators were marginally more likely to disclose to a therapist than those raped by one perpetrator only Disclosures to informal sources of support F (1, 1378) =0.59, p = .84, and involvement
86 of the cri minal justice system F (1, 1378) =0.44, p = .52, were unrelated to victimization experience. Research Question One A logistic regression model addressed research question 1 ) Which ecologically based factors in a model of sexual repeat victimization predic t respondents victimized by one perpetrator only versus repeat victims of sexual violence? Findings indicated moderate support for the included ecological variables with the Hosmer and Lemeshow finding R2 = 0. 17 and the overall model significant 2 (19) = 1 5 0. 36, p = 000 (see Table A 2 ). In t he substantial equation model, t wo control variables assessed the influence of the time gap between the phone interview and rape victimization experience. The recency variable did not significantly influence the odds of repeat victimization5, but an older age during the rape victimization decreased the odds of future victimization by a different perpetrator by 4% (AOR=0.96, se = 0.02, p = .0 1 ). Within the childhood development level, neither having a child before the age of 18 nor experiencing childhood maltreatment predicted repeat rape victimization by different offenders The lack of influence of childhood maltreatment is notable, as the experience of childhood maltreatment was found to increase the odds of any victimization by 2,703% in the Heckman selection model differentiating the odds of nonvictimization versus any rape victimization. 5 Again, repeat victimization does not refer to the num ber of rape incidents but to the number of perpetrators, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
87 In the context of victimization by the first or only perpetrator category victim offe nder relationship was found to significantly predict the odds of repeat rape victimization by distinct offenders Rape victimiz ation by a stranger decreased the odds of repeat victimization by 73% (AOR=0.27, se = 0.43, p = .002). Likewise, rape victimizati on by an acquaintance decreased the odds of repeat victimization by distinct offenders by 46% (AOR=0.54, se = 0.30, p = 0.04), in comparison with referent group of first time victimization by a family member. V ictimization by current or ex -partners was unr elated to repeat rape victimization by a different perpetrator. T he pre victimization context also significantly related to repeat rape victimization by different offenders such that hanging out prior to the rape victimization increased the odds of repeat victimization by a different offender by 1 56% (AOR=2. 56, se = 0.41, p = 0.02). Alternately, drinking or doing drugs being on a date or engaging in sexual activity and fighting prior to the rape victimization were unrelated to repeat rape victimiza tion. These findings contradict expectations that factors that may increase victim blame would predict vulnerability to repeat victimization. Follow up investigation did not indicate that the influence of doing nothing on the odds of repeat victimization was driven by its correlation with other significant predictors of repeat victimizations such as victim -offender relationship, pregnancy as a result of the victimization or age at victimization by the first or only offender Through examining standardi zed residuals in Chi -Square tests, those who were just hanging out were not significantly more likely to become pregnant after victimization by the first or only offender (z = 1.2), or to be victimized by strangers (z = 0 4 ), acquaintances (z = 0.4), pa rtners (z = 1.5), or family (z = 0. 1 ) (See Table A 3) Likewise, those who reported
88 just "hanging out" prior to the rape victimization versus those who were using substances, on a date or engaging in sexual activity, or fighting with the offender did not significantly vary regarding age at time of the victimization, F (1, 1364) = 0.96, p = .59. Both o f t he last two variables in the context of the victimization category significantly related to repeat rape victimization by distinct offenders As offender coercion tactics transitioned from verbal coercion to physical violence or use of a weapon, the odds of repeat victimization increased by 32 % at (AOR=1. 32, se = 0.1 2 p = 0.0 2 ). Unexpectedly, if the victim became pregnant as a result of the assault, the odds of repeat rape victimization by different offenders decreased by 9 0% (AOR= 0.10, se = 1.03, p = 0.02). Further examination to determine if pregnancy outcome related to repeat victimization was unsuccessful as only respondents victimized by one perpetrat or only answered the subsequent question regarding whether the preg nancy resulted in a live birth. Within the social power level, race/ethnicity, income, victim disclosures, and engagement with the criminal justice system were unrelated to the odds of repeat rape victimization by distinct offenders Lastly, lambda was significant (AOR=0.84, se = 0. 04, p = 0.000 ), indicating that the variables contributing to selection into a sample of rape victims are significantly associated with those influencing selectio n into a sample of repeat rape victims. Research Question Two Post -Victimization Mental Health among All Rape Victims : Bivariate Relationships Relationships between predictor variables and post -victimization mental health are first assessed at the bivariate level for all victims. Recency of the rape victimization by the first or only offender = .0 0 t (1379) = 0.86, p = 0.41, and age at the rape
89 victimization by the first or only offender = .00, t (1379) = 0.00, p = 0.80, were unrelated to depression at the time of interview A h istory of chronic health problems, = .21 t (1379) = 7.81, p = 0.000 significantly predicted more severe depres sion scores. Experiencing multiple rape victimizations by separate offenders = 10, t (1379) = 3.54, p = 0.000 also significantly and positively related to depression symptoms on a bivariate level Within the context of the rape victimization by the first or only perpetrator the use of physical violence as a coercion tactic, = .03 t (1379) = 1.30, p = 0.20, and pregnancy as a result of rape victimization = -.03, t (1379) = 1.13, p = 0.26, were unrelated to depression at the time of interview for all victims. Victim -offender relationship F (3,1377) = 2.96, p = 0.03, significantly relate d to depression symptoms at interview. Due to adherence to the assumption of homogeneity of variance, a Tukey post hoc test was selected. Post hoc analysis reve aled the mean depression score for those victimized by partners (M=.40, SD=.21) or family (M=.40, SD=.40), were borderline significantly different at p = .08 than those victimized by acquaintances (M=.37, SD=.18). H anging out, = .01, t (1379) = 0.60, p = 0.55, using substances, = .0 2 t (1379) = 0. 45, p = 0. 6 5, being on a date or engaging in sexual activity, = .02, t (1379) = 1.29, p = 0.20, or fighting with the perpetrator, = .02, t (1379) = 0.26, p = 0.79, prior to the rape victimization by the first or only offender were unrelated to depression on a bivariate level. Time off from school and work after rape victimization by the first or only perpetrator significantly predicted depression, = .11 t (1379) = 4.07, p = .000. Race/ethnicity, = .05 t (1379) = 2.02, p = 0.05, and income, = -.13
90 t (1379) = 4.98, p = 0. 000, were both associated with depression at the time of interview such that being white and having a higher income were protective against depression Likewise higher education levels were protective against depression = 21, t (1379) = 7.80, p = 0.000. Therapist disclosure, = .03 t (1379) = 1.17, p = 0 .24 and victim engagement of the criminal justice system, = .01, t (1379) = 0.55 p = 0 .58 were not signif icantly linked to depression at the time of interview for the full sample of rape victims on a bivariate level. Post -Victimization Mental Health among Respondents Sexually Victimized by One Perpetrator Only: Bivariate Relationships Among respondents victimized by one offender only recency of the rape victimization by that offender, = .01 t (1153) = 0.49 p = 0.63, and age at rape victimization, = .00, t (1153) = 0.77, p = 0.44,were unrelated to post -victimization mental health. A history of health problems = .18 t (1153) = 6.43, p = 0.0 00, predicted more severe depression scores among respondents sexually victimized by one offender only on a bivariate level Within the context of the rape victimization, the use of physical violence as a coercion tactic = .02 t (1153) = 0.55, p = 0.58, and pregnancy, = -.03, t (1153) = 1.14, p = 0.25, did not influence depression. Victim -offender relationship F (3, 1237) = 2.39, p = .07, only marginally related to depression at the time of interview among women raped by one offender at p = .0 7 Hanging out, = .02, t (1153) = 0.61, p = 0.55, using substances, = .02, t (1153) = 0.20, p = 0.84, being on a date or engaging in sexual activity, = .02, t (1153) = 0.75, p = 0.46, and fighting with the perpetrator, = 02, t (1153) = 0.10, p = 0.93, prior to the rape victimization were unrelated to depression
91 for women sexually victimized by one offender only. Time off from school and work after the rape victimization si g nificantly predicted depression, = .11 t (1153) = 3.73, p = .000, on a bivariate level In contrast to bivariate findings for all victims, r ace/ethnicity, = .03 t (1153) = 0.91, p = 0.37, was unrelated to depression scores for women sexually victimized by one offender only. A greater income level was protective against depression = -.13 t (1153) = 4.77, p = 0. 000, similar to education level = -.20, t (1153) = 7.15, p = 0.000. Therapist disclosure, = .01 t (1153) = 0.46, p = 0.65, and victim engagement of the criminal justice system, = .02, t (1153) = 0.84 p = 0. 41, were not significantly linked to depression at the time of interview for respondents sexually vic timized by one perpetrator only on a bivariate level Post -Victimization Mental Health among Respondents Sexually Victimized Multiple Times by Separate Offenders : Bivariate Relationships Among respondents sexually victimized by multiple offenders over time recency of the rape victimization by th e first offen der, = .00 t (138) = 0.03, p = 0.85, and age at victimization = .00, t (138) = 0.77, p = 0.44, were unrelated to long term post victimization mental health. Consistent with one -perpetrator victims, a history of health problems = .36 t (138) = 4.51, p = 0.000, significantly predicted depression at the time of interview for respondents sexually victimized multiple times by distinct offenders. Within the context of the rape victimization by the first offender, the use of physical violence as a coerci on tactic = 0.14, t (138) = 1.61, p = 0 .11 and pregnancy were unconnected to depression for women sexually victimized by separate perpetrators over time, = 0.13, t (138) = 1.49, p = 0.14. Victim -offender relationship F (3, 136) =
92 0.21, p = .89, did not relate to depression at the time of interview among women repeat victimized by separate offenders. Hanging out, = .0 4 t (138) = 0. 76, p = 0. 4 5, using substances, = .0 5 t (138) = 1.08, p = 0. 28, being on a date or engaging in sexual activity, = .0 6 t (138) = 1 62, p = 0. 11, and fighting with the perpetrator, = .0 5 t (138) = 1 40, p = 0. 16, prior to the rape victimization were unrelated to depression on a bivariate level for women repeat vic timized by separate perpetrators A lthough time off from school and work after victimization significantly predicted depression for one perpetrator rape victims, this relationship was not significant among respondents repeat victimized by different offende rs = 0.14, t (138) = 1.61, p = 0.11. In contrast to findings for the one -perpetrator victimization sample, being white protected against depression = 0.25, t (138) = 3.03, p = 0.003, and i ncome, = .10 t (138) = 1.14, p = 0.30, w as no t associated with depression for women repeat victimized by distinct offenders Consistent with findings among one -perpetrator rape victims, higher education levels we re protective against depression, = .26 t (138) = 3.1 7, p = 0.002 for repeat victims Th erapist disclosure, = 13, t (138) = 1.55, p = 0. 1 2, and victim engagement of the criminal justice system, = -.04 t (138) = 0.51 p = 0. 61, were not significantly associated with depression at the time of interview for respondents repeat sexually victi mized by different perpetrators on a bivariate level. Multivariate linear regression assesses research question 2) Which ecologically based factors in a model of longterm post victimization mental health predict depression in respondents victimized by one perpetrator only versus repeat victims6 of sexual 6 Again, repeat victimization does not refer to the number of rape incidents but to the number of perpetrator s, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
93 violence? OLS examines the relationships between each factor in long -term depression while controlling for the influence of all other indic ators. A full model including both respondents victimized by one perpetrator only and repeat victims of rape by separate perpetrators was assessed first, and then respondents victimized by one perpetrator and repeat victims were modeled separately to exami ne the relationships between ecological predictors and depression (See Table A 2 ). Post -Victimization Mental Health among All Rape Victims : Multivariate Relationships In the substantial equation assessing longterm mental health outcomes among all victims continuous and ordinal variables were entered collectively in the first step and victim -offender relationship was entered in the second step. The first step explained a moderate amount of variance (Adjusted R2 = 0.11, F (18,1346 ) = 9.85, p = .000). Due to the lack of impact on victim -offender relationship, the final model was similar to the model in the first step (Adjusted R2 = 0.1 1 F (2 1 ,1343 ) = 8.71, p = .000). In the second step, the control variable recency was not significantly related to long -term depression, = .00, t (1362) = 0.07, p = .95. A mong childhood development variables, both repeat rape victimization by separate offenders = 06, t (1362) = 2.03, p = .0 4 and history of health problems were significantly related to depression, = 1 7 t (1362) = 6. 42, p = .0 00, consistent with bivariate findings. In the context of the victimization by the first or only perpetrator category p regnancy as a result of rape victimization protected against later depression symptoms = .0 6 t (1362) = 2. 3 9 p = 0 2 in contrast to bivariate findings which documented no relationship. T aking t ime off from school and/or work predicted greater depression =
94 10, t (1362) = 3. 4 3 p = .0 0 1 Diverging again from non -significant bivariate find ings, w ithin the social power level, both income, = -.07, t (1362) = 2. 0 8 p = .0 4 and education, = -.1 8 t (1362) = 6. 51, p = .000, were protective against depression. Age at victimization by the first or only offender interaction between the victim and offender prior to the rape victimization, and post victimization support seeking were unrelated to depression. Lambda was significantly related to depression, = -.11, t (1362) = 4.1 7 p = .000, indicating that the var iables contributing to selection into a sample of rape victims are significantly correlated with the variables associated with depression levels in the full sample. Differences between bivariate and multivariate models suggest that some predictors may act as suppressor variables, leading to underestimated relationships when investigated on the bivariate level. Subsequent regression analyses model long -term depression outcome separately for respondents victimized by one perpetrator only versus repeat victims of rape by separate perpetrators revealing variations in predictors of depression. Post -Victimization Mental Health for Respondents Sexually Victimized by One Perpetrator Only : Multivariate Relationships Among respondents sexually victimized by one per petrator the first step of the model explained a small amount of variance (Adjusted R2 = 0.09, F (1 7 ,1208 ) = 8.15, p = .000). Just as in the full model of all victims victim offender relationship did not influence depression for one perpetrator rape victims leaving the final model similar to that in step one, (Adjusted R2 = 0.09, F (2 0 ,1205 ) = 7.12, p = .000). Significant predictors in the final step of the one -perpetrator rape victims model were consistent with those found in the full model examini ng all victims Among
95 childhood development variables only a history of health problems were significantly related to greater depression at the time of interview = .1 5 t (1223) = 5. 2 8 p = 0. 000. In the context of the rape victimization category p regnancy as a result of rape victimization was negatively associated with depression for one perpetrator rape victims = .0 7 t (1223) = 2. 54, p = .0 1 The use of physical violence as a coercion tactic was marginally significantly predictive of depression symptoms at p = .10, = -.05, t (1223) = 1.6 2 T aking t ime off from school and/or work predicted greater depression = .11 t (1223) = 3.3 5 p = .0 0 1 among respondents victimized by one perpetrator only Within the social power level, both education = .1 7 t (1223) = 5. 59, p = .000, and income, = .08, t (1223) = 2. 3 7 p = .02, were protective against depression. Age at sexual victimization, pre -victimization activities, and post-victimization support seeking were unrelated to depressi on for respondents victimized by one perpetrator only. Lambda was significantly related to depression, = -.1 2 t (1223) = 3. 97, p = .000, indicating a selection effect among a sample of women victimized by one perpetrator only. Post -Victimization Mental Health for Respondents Sexually Victimized Multiple Times by Separate Offenders : Multivariate Relationships Among repeat victims of rape by separate offenders several of the effects found in the full model and one -perpetrator rape victim models disappeared, and several unique ef fe cts emerged. Similar to the models for all victims and one -perpetrator rape victims, step one explained most of the variance in the repeat victim7 model (Adjusted R2 = 0.23, 7 Again, repeat victimization does not refer to the number of rape incidents but to the number of perpetrators, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
96 F (1 7 ,12 1 ) = 3.44, p = .000), varying only somewhat from the final model (Adjusted R2 = 0.27, F (2 0 ,11 8 ) = 3.46, p = .000). The final step of the model found a history of health problems were significantly related to depression, = .3 7 t (1 36) = 4.27, p = .0 00, for repeat victi ms Coefficient comparison tests indicate a significant difference based on repeat victimization history regarding the influence of chronic health problems on depression, t (1362 ) = 2.16, p = .03, such that a stronger effect is found among repeat victims. In the context of the rape victimization level, doing nothing prior to the victimization protected against depression at the time of interview = 0 22, t (136) = 2.06, p = 0.0 4 for women victimized by multiple perpetrators over time. In addition, havin g the first offender be a n acquaintance rather than a family member was predictive of longterm depression, = 2 6 t (136) = 2.70, p = 0.0 1, and an offender that was a partner rather than a family member predicted greater depression at p = .09 = 17, t (136) = 1.68. This borderline relationship may reflect a longer average recovery period between victimization by family members and time of interview than those by acquaintance or partners, although recency of was not significant in the multivariate model. Follow up ANOVA analyses on the relationship between recency of victimization and victim -offender relationship finds the average number of years since the victimization significantly varies, F (3, 136) = 3.38, p = .02. Due to the assumption of homogeneity of variances being violated, Games -Howell post hoc tests were conducted. Recency of victimization by first offenders who were f amily members (M=29.84, SD=13.25) was significantly greater than that by acquaintance s (M=22.0, SD=11.94) and borderline greater than that by strangers at p = .08.
97 In the social power category education, = -.2 5 t (136) = 2.83, p = 0.0 0 5 and police engagement, = -. 17, t (136) = 2.00, p = 0.0 5 were protective against negative mental health outcome among those repeat victimized by multiple perpetrators over time Coefficient comparison tests reveal no significant difference in the influence of education on depression between one -perpetrator victims and repeat victims, t (1362) = 1.1 4 p =.26. Being a woman of color increased longterm depression symptoms = -.21, t (136) = 2.62, p = 0.009, for repeat victims of rape by separate offenders. Lambda was significantly related to depression, = -.16, t (136) = 1.98, p = .05, indicating a selection effect in the sample of repeat victims of rape by distinct offenders. Summary The current research contributes to the body of literature on sexual repeat vi ctimization and post -victimization trauma through employing model s in tegrating ecological individual, situational, and sociocultural variables in a national probability sample. Multivariate logit regression finds that repeat sexual victimization is more likely in respondents when victimization by the first or only offender occurs early in life and is perpetrated by family members with greater physical force in ambiguous pre victimization contexts. In addition, unique risk factors for long term depression in respondents victimized by one perpetrator only versus repeat victims of rape by separate perpetrators a re identified. Multivariate OLS regression indicates in a full model of rape victims that a history of health problems, sexual repeat victimization, and taking time off from work or school after the victimization by the first or only offender predicted greater depression. Higher education, income, and pregnancy as a result of the victimization by the first or only offender were protective against longterm depression symptoms.
98 Findings in a depression model among women se xually victimized by one perpetrator only determine similar risk and protective factors as those in the full model of all victims with the addition of physical violence as a coercion tactic marginally protecting against depression at p < .10 Risk and protective factors in the multivariate model of long-term depression symptoms for repeat victims of rape by separate offenders diverge from those of the full and one -perpetrator rape victim models. Among repeat rape victims, greater quantity of informal suppo rt after the victimization by the first offender predicted greater depression at the time of interview Rape by an acquaintance rather than a family member predicted depression symptoms and rape by a current or ex -partner was marginally associated with de pression symptoms at a p < .10 level. A pre -victimization context of "doing nothing" was protective against depression at the time of interview for respondents repeat victimized by different perpetrators Lastly, being white having higher education levels and police engagement were protective against depression for repeat rape victims These results are considered with respect to extant research findings and policy implications in Chapter 7.
99 CHAPTER VII DISCUSSION Research Question One Findings in the current study suggest vulnerability to repeat victimization increases when rape victimization occurs at a younger age and is perpetrated by family members through physical coercion with little pre victimization warning. Results within each ecological category are considered with respect to previous research. Control Variables Age at victimization by the first or only perpetrator was significantly younger among repeat victims of rape than respondents victimized by one perpetrator only in both biva riate and multivariate analys es consistent with prior research (Casey & Nurius, 2005). Descriptive analyses reveal the average age of rape victimization by the first or only perpetrator was 18 years old for respondents victimized by one offender and less than 15 years old for repeat rape victims8. Childhood D evelopment Within the childhood development category, neither having a child before the age of 18 nor experiencing childhood maltreatment predicted repeat rape victimization by separate offenders in the multivariate model. The absence of a relationship between childhood maltreatment and repeat victimization is unexpected. E xperienc ing childhood maltreatment increase d the odds of any rape victimization by 2,703% in the Heckman selection model, and reports of childhood maltreatment were significantly greater among 8 Again, repeat victimization does not refer to the number of rape incidents but to the number of perpetrators, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
100 repeat rape victims than respondents victimized by one offender only in the bivariate model. Further, Casey and Nurius (2005) found that repeat sexual assault victims had greater nonsexual trauma in childhood than women victimized by one offender only. However, the lack of association between childhood maltreatment and sexual repeat victimization is consistent with Merrill et al. (1999)s analysis, indicating that once the effects of childhood sexual assault were considered childhood physical abuse no longer predicted adult sexual assault The current research findings consistency with one of two previous studies may be result of dependent variable operationalization. Although the sampling method in the current study comports more closely to Ca sey and Nuriuss (2005) general population sample than Merrill et al. s (1999) female Navy recruit sample, Merrill et al. (1999) operationalized adult sexual victimization as rap e, consistent with the current study, whereas Casey and Nurius (2005) measured adult sexual assault. Context of Rape Victimization In the context of the rape victimization by the first or only perpetrator victim offender relationship significantly influe nced the likelihood of repeat rape victimization by separate offenders in both bivariate and multivariate models. In comparison to respondents whose first or only assailants were family members, the odds of repeat victimization were 73% lower for respondents victimized by strangers, and 63% lower for respondents victimized by an acquaintance. Follow up analyses reveal ed p rimary rape victimization by a family member related to age at victimization and by proxy, lifetime exposure to perpetrators Respondents whose first or only rape victimization was perpetrated by a family member were on average minors during the victi mization (M = 9.9 5 SD = 6.19), whereas respondents victimized by acquaintances (M = 18.4 3 SD =
101 8.06), strangers (M = 18. 86, SD = 8.45), or current or former partners (M = 24.17, SD = 9.37) were more likely to be adults at the time of victimization, F (3, 1380) = 154.23, p = .000. In addition, with the literature documenting more negative reactions to disclosures of sexual victimization by family members (Ullman, 2007), it is possible negative social support may additionally influence the relationship betwe en victim -offender relationship and sexual repeat victimization. P re -victimization context significantly related to repeat rape victi mization by separate offenders in both bivariate and multivariate analyses Respondents who reported doing nothing prior to the first or only rape victimization had a 156% increased odds of repeat victimization. Findings revealing that victim behavior is unrelated to repeat rape victimization depart from previous research linking victim alcohol use and sexual behavior to repeat victimization (Fargo, 2009; Greene & Navarro, 1998; Himelein, 1995; Krah et al., 1999; Krebs et al., 2009; Fisher et al., 2000; Testa et al., 2010). However, pre victimization behaviors represent an isolated incident rather than a gen eral pattern of substance use A follow up ANOVA assessed responses to three questions regarding a) frequency of alcohol use over the previous 12 months, b) frequency of alcohol use over the previous two weeks, and c) average number of drinks per day in the previous two weeks. Respondents victimized by one perpetrator only and repeat victims did not differ with respect to frequency of alcohol use over the past year ( F (1, 1379) = .00, p = 0.98) or two weeks ( F (1, 890) = .29, p = 0.59). H owever, among respondents reporting drinking in the previous two weeks, repeat victims (M = 4.50, SD = 5.69) consumed a significantly higher number of drinks than respondents victimized by one perpetrator only (M = 2.43, SD = 2.25) F (1, 492) = 23.70, p = 000. Again, one is unable to
102 extrapolate an ongoing pattern of substance use based solely on substance use at the time of interview although research documents the heritability of substance use at bet ween 40% and 50% (Goldman, Koss, & Russ, 2005). Result s demonstrating that being on a date or engaging in sexual activity were unrelated to repeat rape victimization by distinct offenders dispute s research connecting frequent consensual sexual activity with repeat victimization ( Fargo, 2009; Himelein, 1995; Krah et al. 1999), and replicates studies finding a minimal or no relationship between sexual behavior and repeat victimization (Mayall & Gold, 1995; Merrill et al., 1999; West et al., 2000). However, similar to alcohol use it is problematic to assume p re victimization behavior characterizes general pattern s of sexual activity Although the Tjaden and Thoennes (1998) survey did not comprehensively measure respondents sexual histories one section included items on cohabitation and marriage, as well as other sexual involvement s at the time of interview ANOVA tests found repeat victimization related to number of marriages where among current or previously married respondents repeat victims (M = 1.25, SD = 2.01) reported a higher average number of marri ages than did one -perpetrator rape victims (M = 0.84, SD = 1.84), F (1, 1114) = 5.46, p < .05). Quantity of intimate partnerships is not a direct measure of sexual behavior, but Wyatt et al. (1992) documented an association between multiple partnerships short sexual relationships and repeat victimization. However, the current study finds only one measure of relationship history to associate with repeat victimization, leaving the association between patterns of intimacy and repeat victimization inconclus ive. Additionally, the finding that hanging out or doing nothing in particular predicted repeat rape victimization may support the literature on poor perception of threat
103 and sexual repeat victimization by separate offenders (Katz et al., 2010). Experi encing a rape victimization precipitated by ordinary circumstances may impair survivors perception of threat, increasing the odds of rape victimization by a new offender Studies documenting indirect relationships between threat perception and repeat vict imization (Katz et al. 2010; Macy, Nurius, & Norris, 2006; Nurius et al. 2000) suggest future research should consider pre -victimization context as a moderator in the relationship between threat perception and sexual repeat victimization The remaining two variables in the context of rape victimization category both significantly related to sexual repeat victimization in bivariate and multivariate models. Offender coercion tactic s of physical violence or the use of a weapon increased the odds of repeat victimization by a different offender by 30%, consistent with Griffin and Reads (2012) and Casey and Nuriuss (2005) findings. Unexpectedly, if respondents be came pregnant as a result of rape by the first or only offender the odds of repeat rape victimiz ation decreased by 90% Follow up analyses on pregnancy outcome were unsuccessful, as only victims of rape by one perpetrator reported whether the pregnancy resulted in a live birth. Previous research identifies unwanted pregnancies and abortions as risk f actors for sexual repeat victimization in a community sample of women (Wyatt et al., 1992), although pregnancies resulting from rape victimization were not exclusively considered Alternately, Collins (1998) documented in a longitudinal sample of adolescen t mothers th at pregnancy at T ime 1 reduced likelihood of sexual assault at T ime 2 Collins (1998) likewise did not anticipate this finding but suggested pregnancy may motivate offenders to select other targets, or reduce women's exposure to offenders through increasingly constrained behavior as the pregnancy progresses.
104 Social P ower S ocial power variables including race/ethnicity, income, victim disclosures, and engagement of the criminal justice system were all unrelated to the odds of repeat rape victimization in the multivariate model The absence of a relationship between race/ethnicity and repeat victimization contrasts Urquiza and Goodlin Jones (1994) findings, although the current sample was far more racially and ethnically homogenous than Ur quiza and Goodlin Jones (1994) sample, restricting comparison Contrary to the literature associating low socio-economic status and education levels with sexual and interpersonal repeat victimization ( Acierno et al., 1997; Kimerling, et al. 2007; Mears e t al., 2001) the present study found no such relationships This may be due to the complicated relationship between sexual victimization higher education, and income Although higher education increases average lifetime earnings (Carnevale, Rose, & Cheah 2011) individuals attending institutions of higher education experience greater relative risk of sexual repeat victimization (Kilpatrick et al., 1998). In the current sample each level of educational attainment increased the odds of any sexual victimization by eight percent in the Heckman selection model As a result, although economic privilege may facilitate survivor access to systems of care and justice, the pathway to higher incomes, higher education, conversely increases women's vulnerabili ty to sexual repeat victimization. Disclosures of the rape victimization by the first or only offender to informal sources of support or therapist s were unrelated to repeat rape victimization by separate offenders These findings are congruent with resear ch determining that the act of disclosure alone does not influence likelihood of repeat victimization by a new offender
105 (Filipas & Ullman, 2006). In fact, maladaptive coping strategies of substance use, social isolation, and acting out sexually or aggres sively appear to affect the odds of repeat victimization more than adaptive coping strategies of disclosing to social support networks or therapists (Filipas & Ullman, 2006). V ictim engagement of the criminal justice system was unrelated to the odds of rep eat rape victimization potentially due to low prosecution rates of sexual assault cases (United States Department of Justice, 2010) and/or experiences with criminal justice professionals that left victims feeling blamed or at fault for their victimization (Ullman & Filipas, 2001). However follow up analyses among women reporting filing a police report revealed police satisfaction on a four -point scale from very satisfied (1) to very dissatisfied (4) did not significantly vary between wo men victimized by one offender (M = 2.56, SD = 1.18) and repeat victims (M = 2.63, SD = 1.16) Research findings in the current study support ecological assumptions that the origins of sexual repeat victimization extend beyond victim history and behavior, as reports of childhood maltreatment and pre -victimization drinking, sexual activity, and/ or fighting with the perpetrator were unrelated to the odds of repeat victimization by a new offender In fact, likelihood of repeat rape victimization escalated when the first or only offender targeted younger victims or family members, used physical force, and/or exploited innocuous, ordinary settings to perpetrate their crime The odds of repeat victimization were unaffected by t he number of formal and informal sour ces of support respondents confided in replicating prior research determining the quality of the reaction to a disclosure is more consequential than the act of disclosure itself.
106 Research Question Two Informed by research finding repeat victims suffer more serious mental health symptomology than those victimized by one perpetrator only (Banyard et al. 2001), research question two developed and assessed a model of post -victimization depression to examine separately the mental health trajecto ries of women victimized by one offender versus repeat victims Campbell et al.s (2009) position that post -victimization trauma emerges not only in response to the victimization itself, but also to socio -cultural treatment of victims guided the current study to invest igate whether variables beyond sexual victimization history predict depression among survivors Results identified unique risk and protective factors for depression among women victimized by one perpetrator only and repeat victims of rape. Control Variab les The number of years since the rape victimization by the first or only perpetrator occurred did not influence long -term depression in the bivariate or multivariate analyses irrespective of repeat victimization history These findings are duplicative, as one meta analysis of 37 studies on child sexual abuse found a similar measure, age at victimization, unrelated to a range of negative psychological and behavioral outcomes (Paolucci, Genius, & Violato, 2001). Childhood D evelopment Consistent with previ ous research, repeat rape victimization by separate perpetrators wa s positively associated with greater depression (Arata, 2000; Banyard et al. 2001; Kimerling et al. 2007; Messman-Moore et al. 2000). As expected, h istory of health problems predicted lo ngterm depression for all victim types, replicating research
107 associating sexual victimization with poor physical and mental health (Cloutier et al., 2002). The extent of the significant relationship between physical and mental health among victims was inf luenced by the extent of sexual trauma. Coefficient comparison tests determined the relationship between health problems and depression was stronger among repeat victims than for respondents victimized by one perpetrator only. Cloutier et al. (2002) simila rly connected severity of sexual victimization experiences with physical health problems, finding hypertension, high cholesterol, diabetes, and obesity more likely among victims of rape than victims of sexual assault (Cloutier et al., 2002). Context of Rape Victimization Variables Several indicators of depression varied between respondents victimized by one perpetrator and repeat victims. D oing nothing prior to the rape victimization by the first perpetrator acted as a protective factor against long term depression for repeat rape victims only. This finding may suggest that conformity to stereotypical rape scripts reduces the risk of longterm depression among repeat rape victims, as attributions of victim responsibility for a sexual assault ar e predicted by the belief that victim behavior encouraged the perpetrator (Dietz, Littman, & Bentley, 1984). The inconsequence of substance use prior to the victimization on psychological outcome corresponds with the limited research assessing this relationship (Campbell et al., 1999; Resnick et al., 1997). Similarly, in the rape victimization by one perpetrator only model, the use of physical violence as a coercion tactic wa s marginally protective against depression at p = 10. Again, this may imply conf ormity to stereotypical rape scripts involving a physically aggressive offender is protective against depression for respondents sexually victimized by one perpetrator only, but the relationship is
108 insubstantial. These findings may be weak as a result of m ediation by social reactions to disclosures that result in greater trauma (Orchowski et al., in press ; Ullman et al., 2006), or self -blame (Ullman, 2007). Victim offender relationship was found to be unrelated to depression for women victimized by one off ender only replicating some findings (Campbell et al., 1999; Kramer & Green, 1991; Mackey et al., 1992; Riggs, Kilpatrick, & Resnick, 1992), and contradicting other studies relating either stranger or known offenders assaults with greater depression and/or trauma (Bownes et al., 1991; Culbertson & Dehle, 2001; Ellis, Atkeson, & Calhoun, 1981; Masho & Ahmed, 2007; Ullman et al., 2006) In the model assessing long term depression among repeat victims primary victimization by a family member rather than an acquaintance was unexpectedly protective. Follow up analysis indicated time since victimization was significantly greater for those victimized by family than acquaintances or partners However, recency was not linked to lo ngterm mental h ealth in the m ultivariate model, consistent with some previous research (Paolucci et al., 2001 For respondents victimized by one perpetrator only pregnancy protected against long -term depression and taking time off from school and/or work were risk factors for greater depression. Further investigation into pregnancy outcome among respondents victimized by one offender only revealed that sixty percent of respondents impregnated through rape report ed giving a live birth, but it is unknown whether respondents raised the in fants. The association between time off and post -victimization mental health is consistent with previous research finding avoidance behavior associated with higher levels of depression post victimization (Frazier et al. 2005; Gutner et al. 2006; Ullman,
109 et al., 2007). A voidance coping may also be a response to negative social support which is another risk factor for post -victimization distress (Ullman et al. 2007, Ullman, 1996c). Social P ower Within the social power category education wa s protective a gainst longterm depression for both victim types Income was additionally protective against depression for respondents victimized by one perpetrator only As Campbell et al. (2009) note, the relationship between education level and post -victimization dis tress is unclear In the present study, education appears protective against depression, despite increasing the odds of becoming a rape victim versus a non -victim by 8% with each subsequent step in education. However, education serving as a protective fact or is consistent with the findings of another random -digit -dialing sample examining the impact of sexual assault and rape, although PTSD, rather than depression was the outcome variable (Masho & Ahmed, 2007) Race/ethnicity was unrelated to depression among women victimized by one perpetrator only, consistent with previous research (Ahrens et al. 2010; Elliott et al., 2004; Frank & Stewart, 1984; McFarlane et al., 2005; Sorensen & Siegel, 2010; Wyatt, 1992). However, being white was a protective factor for respondents repeat victimized by separate offenders indicating a potential interaction between race/ethnicity and rape victimization history on mental health outcome. Police engagement shielded against long -term depression for repeat rape victims, co nsistent with Masho and Ahmeds (2007) findings that talking with police predicted fewer PTSD symptoms Results indicated informal support acted as a risk factor for depression among repeat rape victims only, possibly revealing differential experiences wit h disclosures of rape victimization between
110 women victimized by one offender and repeat rape victims Johnson and Johnson ( in press ) identified high quality social support as a protective moderated mediator for the relationship between severe sexual trauma and problematic substance use Conversely, a greater quantity of social support acted as a risk moderated mediator (Johnson & Johnson, in press ). P olice engagement guard ed against long term depression for survivors of repeat rape victimization indicating access to justice after victimization by the first offender may be of greater relative importance to longterm recovery among repeat victims The overall model for women victimized by one offender only significantly predicted depression, but explained l ittle of the variance Alternately, the repeat rape victimization model accounted for a moderate amount of variance in psychological outcome. Model weakness may be due to the retrospective nature of the methods. In spite of these limitations the current res earch contributes to our understanding of the greater mental health issues faced by repeat victims. With variables other than repeat rape victimization predicting mental health outcome, support is provided for the notion that victimization trauma emerges f rom socio -cultural factors beyond the victim and victimization itself. Research Limitations Although the large sample size and diverse demographics are strengths of the data used in the current study, there are serious methodological limitations. The data are cross sectional, curtailing conclusions regarding temporal order. The retrospective nature of the research is problematic, with the average length between time of interview and victimization by the first or only offender at around twenty years. One recent review of
111 research on the accuracy of memories of traumatic events finds that for single, time limited and ongoing t ra u matic events, memories are generally correct not dissimilar to memories for non t ra u matic events (Pezdek & Taylor, 2002). Ho wever, over time, accuracy of recall diminishes and with a n average 20 years between the interview and rape victimization, there are legitimate concerns regarding the accuracy of the findings (Pezdek & Taylor, 2002). In addition although the present data set yielded 1,381 total victims, only 140 of those victims were repeat victims, hindering multivariate analy tic power Another limitation of the present data is the inability to develop an ordinal measur e of sexual victimization Ideally, a measure of victimization history would include one time victimization by one perpetrator, multiple victimizations by one perpetrator, and multiple victimizations by distinct perpetrators. How ever, the nature of the measure used in the present study precludes differe ntiation beyond comparing women victimized by one offender with women victimized by multiple offenders In addition, rape i s measured in the NVAWS, rather than a continuum of sexual violence that ranges from noncontact sexual aggression to rape, such as th e Sexual Experiences Survey (Koss et al., 2006), commonly used in sexual repeat victimization research. However, some research suggests there is not an apparent continuum of trauma severity in sexual victimization. Testa et al. (2004) found limited support for an ordinal sexual victimization measure, where rape was associated with greater trauma than verbal coercion, but other types of unwanted sexual experiences had similar trauma levels. Likewise, Clum, Nishith, and Calhoun (2002) found that s urvivors of rape by physical or substance coercion did not differ in their reports of victimization severity and although Brown et
112 al. s (2009) respondents generally rated verbal coercion, incapacitated rape, and forcible rape from least to most traumatic, this order disappeared in some domains, such as current perceived trauma and emotional impact. Irrespective of whether an ordinal measure of sexual victimization accurately represents a trauma construct access to information on a range of sexual victimization exper iences would provide greater variance in the dependent variable. External validity is problematic in this study due to selection bias. The response rate among women was 72.1%, limiting generalization of statistical findings Women who agree d to participate in the NVAWS may be unrepresentative of general population, particularly due to the graphic nature of the questionnaire. The study has a diverse large sample and a fairly high response rate, but the NVAWS does not weight the data to account for differenc es between those who participated and those who declined to participate In addition, the research was restricted to households with telephones, excluding homeless and transient populations and households without a working phone line. Yet homeless women are exceptionally vulnerable to sexual assault Wenzel and colleagues found that over 10% of homeless women reported rape victimization in the previous 12 months, and almost 10% experienced sexual assault in the previous month (Wenzel, Koegel & Gelb erg, 2000; Wenzel, Leake, & Gelberg, 2000). In addition, for the post victimization long term depression model, although the average length of time between the victimization by the first or only perpetrator and the interview was almost 20 years, n ine respo ndents reported their victimization occurred the year of the interview, leaving the time order of the first victimization and onset of depression symptoms questionable. Lastly the data were collected 14 to 16 years ago, again threatening
113 external validity However, the age of these data are not necessarily a complication for the current studys intent as the direction of the relationships among variables relating to power and inequality, such as gender, race, and socioeconomic class are stable (DeNavas-W alt et al. 2009; Shaefer & Edin, 2013; U.S. Bureau of the Census, 2011). In spite of methodological limitations, the present study identified risk and protective factors within multivariate models of sexual repeat victimization and post victimization depression contributing to the first step of theory building. Although the data in the current study are retrospective and cross -sectional, they provided a unique opportunity to assess a rare populati on: repeat rape victims. Sexual repeat victimization research remains in early stages of theory development and the present study responds to recent scholarly suggestions to situate sexual repeat victimization within an ecological perspective Subsequent research replicating the current research should employ a prospective method to identify risk and protective factors with greater accuracy and determine the causal mechanisms driving relationships between predictors, repeat victimization, and post victimiz ation mental health (Marx et al., 2005). Future Research Future research on ecological predictors of sexual repeat victimization and post victimization depression will respond to limitations in the current research. Although the sample size of Tjaden and Thoennes (1998) data is ideal, the data are over 10 years old, and missing data issues prevent the inclusion of some variables of interest. The National Institute of Justice is currently replicating the NVAWS and these new data may offer several advantag es. First, the demographic makeup of American population has markedly changed in the past decade (U.S. Census, 1992a, 1992b, 2011). Almost n inety percent of
114 the current studys sample was W hite or African American or Black resulting in the creation of a d ichotomous race/ethnicity variable Second, socio -structural variables such as race, income, or education relate to power and inequality across data collection time periods. S hould both NVAWS surveys use the same measures for constructs of interest the da ta set could be collapsed to include both time periods Consequently, multivariate models could be reanalyzed with greater power in a larger data set with collection data as a control variable. Ultimately, a rigorous assessment of ecological predictors of repeat victimization and post -victimization mental health requires a prospective methodology to isolate causes from effects and variable interactions Victimization research is primarily based on cross sectional, large data sets due to the rare nature o f the phenomenon. A prospective examination of sexual repeat victimization and post assault distress that integrates quantitative and qualitative measures has the potential to offer information on risk and protective factors with greater nuance than statis tical association alone Most importantly, a mixed -methods prospective approach would permit closer examination of the causal pathways between risk factors and outcomes, with special attention to potential moderators, mediators or interactions. Isolation of the causal mechanisms relating risk factors to repeat victimization or post -victimization depression additionally assists development of parsimonious model s necessary for quantitative analysis of a phenomenon as rare as repeat victimization. Prospective replication of the current study w ould additionally address a common weakness of secondary data analysis construct validity. M easures selected during primary data collection are based on particular research questions similar but not
115 equivalent to research questions driving secondary data analysis F ew indicators in the NVAWS match the sensitivity of scales commonly used in sexual repeat victimization research to measure constructs such as cumulative trauma, self -efficacy, social reactions to dis closures, coping strategies, and perceived cultural attributions. Alternative Validated Measures A central limitation of the current studies ecological assessment of sexual repeat victimization and post -victimization depression regards the lack of valida ted measures of constructs of i nterest, such as social support, coping strategies and victim self -blame Although the NVAWS provide s information on the number of individuals survivors confided in research determined that quality of social support is more relevant to post sexual victimization recovery than quantity (Johnson & Johnson, in press ). One commonly used measure on social reactions is the Social Reactions Questionnaire (Ullman, 2000) This measure evaluates positive and negative reactions to disclosures of victimization including providing tangible or emotional support, blaming the victim, treating the victim differently, discouraging the victim from disclosing, and reacting egocentrically Similarly the NVAWS survey includes a range of post -victimization questions but there is not a validated measure of coping strategies. One option is the "How I Deal with Things" ( HIDWT: Burt & Kurtz, 1987) five -factor measure assessing cognitive, expressive, nervous/anxious, avoidance, or self -destruc tive coping strategies. Alternately, Ullmans (1996c) coping strategies measure assesses whether survivors used substances, withdrew socially, acted out sexually, sought help through talking about the victimization, consulted a therapist, behaved physicall y aggressively, tried to forget about the victimization, or coped in some other way after a victimization Although the
116 NVAWS data revealed some respondents did not file a police report because they felt culpable for their victimization there was not a st andalone attribution of blame measure Ullman (1996d) measures attributions of responsibility by asking respondents "Thinking back on this experience, how much do you feel each of the following are to blame for your experience?" Response options include : society, offender, own behavior, own personal character, other people, and other factors (Ullman, 1996d) Alternately Meyer a nd Taylor (1986) developed a 15 item attribution statement scale with three clusters of question categories: poor judgment (e.g. "I am a poor judge of character") societal reasons (e.g. "Men have too little respect for women") and victim type (e.g. "I have bad luck"). Both measures assess behavioral and characterological self -blame. Future replications of the current research should integrate validated measures for social reactions to disclosures coping strategies, and self -blame to identify constructs of interest with greater sensitivity and accuracy than the current study. In addition to future research integrating validated measu res to assess the ecological model of sexual repeat victimization scholars should consider theoretical comparison. Weak findings in the current research suggest an ecological model may be inappropriate for explaining the phenomenon of sexual repeat victimization. A priori selection and design of measures to conduct theoretical comparison between ecological and other sexu al repeat victimization models may reveal another perspective has greater explanatory power. Sexual Repeat Victimization Model Comparison Research A possible comparison of repeat sexual victimization models c ould include variables measuring traumatic sexualization, attribution style, tension reduction, routine
117 activities and ecological models. T raumatic sexualization theory (Finkelhor & B rowne, 1985, 1988) proposes that early trauma results in stigmatization, betrayal, powerlessness, and traumatic sexualization Stigmatization causes social withdrawal and/or substance misuse, and betrayal hinders healthy trust and intimacy skills, where one's ability to detect if someone is trustworthy or develop intimacy is compromised. Traumatic sexualization manifests itself as precocious or dysfunctional sexual behavior due to early sexual victimization. Measures of traumatic sexualization theory asse ss self -efficacy feelings of powerlessness age of first consensual sexual behavior risky or avoidant sexual behavior measures of social trust and sensitivity to social contracts, and coping style s of isolation and/or substance use. One common measure o f reactions to social contract norms includes a Wason Selection Task (Stone, Cosmides, Tooby, Kroll, & Knight, 2002), where respondents are presented with if then reasoning rules where one person receives a benefit from another person based on meeting a so cial requirement. Coping strategies could be measured either by the HIDWT scale (Burt & Kurtz, 1987) or Ullmans (1996c) set of questions on post -victimization substance use, social withdrawal, sexual behaviors, help seeking, therapeutic consultation, agg ressive behavior, and attempts to forget the victimization. I nterpersonal agency may be measured using Smith, Kohns, Savage -Stevens, Finch, Ingate, and Lims (2000) scale, which includes five questions on the manner in which respondents interact with other s to achieve particular outcomes (e.g. I accomplish my goals by letting others know my needs and wants"). Peterson and Seligmans (1983) attribution style model suggests that internal, global, and stable attribution s regarding a traumatic event result in learned helplessness and insufficient aggression in response to a threat Measures of the attribution model
118 constitute indicators of attribution style, self -efficacy learned helplessness, threat appraisals and behavioral responses to a potential threat, consistent with those used in Nurius et al. (2000). Threat appraisal involves assessing whether a situation with another person is neutral, dangerous, or beneficial As Nurius et al. (2000) explain, targets of sexual violence commonly confront ambiguous i nformation, such as whether a man's suggestions to leave a social setting together are an appealing move toward intimacy, or an indicator of potential danger. Frequently women balance safety concerns, the fear of social repercussions for misinterpreting a male's sexual or violent interest and the desire to maintain relationships where diplomatic resistance strategies may be elected in order to avoid tension and maintain a friendly rapport with an aggressive male (Nurius et al., 2000) A tension reduction model proposes that sexual behavior and substance use perform as coping strategies in response to PTSD symptomology. This theory could be assessed through a measure of PTSD symptoms and risky sexual behavior, such as the trauma symptom inventory (TSI: Briere, 1995) The TSI measures intrusive experiences, defensive avoidance, anxious arousal, sexual concerns, and dysfunctional sexual behavior, or sexual behavior that is indiscriminate, used for nonsexual goals, and/ or may result in self -harm. In terms o f measuring alcohol use, the drinking habits questionnaire (Cahalan, Cisin, & Crossley, 1969) assesses patterns of alcohol use over the previous month. Lifestyle/routine activities theory (Hindelang, Gottfredson, & Garofalo, 1978) focuses on routine behavi ors of respondents that heighten their risk of repeat victimization. An examination of this theory would assess exposure to crime, or
119 socializing or working in male -dominated environments, victim attractiveness, or frequent substance use, and lack of guard ianship, or traveling alone, consistent with measures used by Fisher et al. (2009). Lastly, an ecological model of repeat victimization would include a consideration of childhood development, victimization context, and social power indicators consistent wi th the current study. In addition to integrating validated measures of social reactions, coping strategies, and self -blame to the current study's methodology, the current study could be extended through incorporating a measure of sexual assault prosecution rates Research indicates the criminal justice system emphasizes prosecution of stranger rapes (Frohmann, 1991) and avoids prosecuting crimes that have been committed against groups with little power, such as prostitutes or ethnic minorities. Indicators o f policies penalizing offenders and prioritizing victim safety include prosecution rates and funds allotted to victim services With the inclusion of measures from each model of sexual repeat victimization a time -series design would comprehensively assess each indicators relationship to sexual repeat victimization and compare the goodness of fit of cluster s of variables representing each model of sexual repeat victimization Because prospective designs are rare, it is valuable to avoid selecting one theory or framework to guide and consequently bias the design, implementation, and interpretation of a study. Future victimization studies employing t heoretical comparison would preclude formulating overly narrow research questions and inadvertently tr uncating variables of interest.
120 Policy Implications Ultimately, findings from the current studies and future research on contextual factors contributing to sexual repeat victimization and post -victimization trauma have the potential to impact policy. Find ings based on an ecological framework suggest a public health approach to addressing sexual violence. A public health intervention involves ongoing systematic collection, analysis, and interpretation of data on the incidence, scope, and risk factors of the phenomenon of interest, in this case sexual repeat victimization and post -victimization trauma This research -based strategy is similar to those which have explained and controlled many communicable diseases. In response to findings from ongoing research, policy development would involve three levels: primary prevention, and secondary and tertiary programs. P rimary prevention would implement education to stop sexually abusive behaviors before they start and raise awareness on post -sexual victimization trau ma. S econdary programs could target individuals at high risk for offending or victimization, prevent first time sexual offenders from recidivating protect victims from repeat victimization and aid survivors in post -victimization recovery. Lastly tertiary programs could stop future violence by offenders with a history of perpetration through community or clinical treatment and monitoring. In addition to the advantages of integrating and prioritizing prevention alongside of punishment, a public health polic y approach shifts the cultural viewpoint on sexual victimization risk from the scale of individuals to the perspective of the population as a whole, on collective rather than individual risk ( Janus, 2006). Based on the findings from the current study, pri mary prevention and secondary programs could focus on several key factors. Results suggest that vulnerability to repeat
121 victimization9 escalates when victimization by the first or only offender occurs at an earlier life stage, and is perpetrated by family members using physical force with little pre victimization warning. Post rape victimization depression was found to be exacerbated by repeat victimization and time off from work or school and all eviated by education and income in the full model In additi on, conformity to stereotypical rape experiences appeared to protect against long -term depression for survivors of victimization by one offender only and repeat victim models Lastly, being W hite and greater police engagement shielded against depression for sexual repeat victims. Findings indicating primary victimization by a family member predict repeat victimization by a new offender directs attention to the problem of familial sexual abuse. From a primary prevention perspec tive, education is central Currently, the United States lacks standardized comprehensive sex education, or a curriculum addressing abstinence, safe sex, and healthy intimacy. As a result, many youth lack the opportunity to learn about the dynamics of sexu al abuse and subsequently disclose any abuse to an adult outside their family. Consistent with previous research, the current study finds sexual violence is targeted toward the young, with the average age of victimization by the first or only offender at 1 8 years of age among women victimized by one offender and less than 15 years old for repeat victi ms. As a result, K 12 curricula on health and relationships should integrate ageappropriate information on sexual violence based on each age groups common vulnerabilities. For example, all age ranges should learn that most off enders are familiar to victims and that verbal coercion is more common than 9 Again, repea t victimization does not refer to the number of rape incidents but to the number of perpetrators, consistent with prior operationalizations of repeat victimization (Casey & Nurius, 2005).
122 physical violence coercion in sexual assault (Fisher et al. 2008). As youth become adolescents, discussions o n drug -or alcohol -facilitated rape will become relevant, as well as what elements characterize healthy, non -violent intimacy. When discussing substance facilitated sexual victimization, the bystander effect should be integrated into curriculum, as social settings with in toxicated males and females place individuals at risk for sexual assault above and beyond use of alcohol (Testa at all, 2010), possibly due to individualistic norms discouraging intervention even when witnessing possibly non consensual sexual behaviors. In the present study, ambiguous pre -victimization contexts predicted sexual repeat victimization by different offenders Similarly, conformity to stereotypical rape script s of a victim doing nothing to provoke the offender was protective ag ainst longterm depression for repeat victims and a physically forceful offender w as borderline significantly protective against depression for women victimized by one offender only Accordingly, a public health campaign addressing attitudes linking victim behavior and responsibility could be an effective component of a primary prevention intervention strategy. Primary prevention should additionally emphasize that the reality of sexual violence is rarely consistent with the stereotype of stranger rape (Ca mpbell et al., 2009). In the current sample, only 14% reported their first or only rape victimization was committed by a stranger, in contrast with 86% who were attacked by current or former partners, acquaintances, or family members. When assessing victim -offender relationship based on rape victimization history repeat victims were much more likely to be assaulted by known offenders (93% versus 85%) and less likely to be assaulted by strangers (7% vs. 15%) than those victimized by one offender only Discl osing victims are commonly
123 expected to isolate which behaviors precipitated the victimization and responses from informal and formal sources of support may be improved by an educational public health campaign. One model is the Center for Disease Control' s Rape Prevention and Education Program, with guiding principles involving educating youth about healthy relationships and changing social norms, both crucial components in addressing interpersonal violence. A p ublic health media intervention about myths r egarding sexual victimization has great potential, with exemplars such as the seatbelt "Click it or ticket" campaign in the United States and the condom use campaign in Brazil establishing that regular and private behaviors can be broadly influenced. Strat egies for secondary programs on sexual repeat victimization are more complex. Advocates for restorative justice reflect on complications the justice system faces when addressing crimes between loved ones, where survivors are unwilling to subject intimate and family abusers to potential incarceration, sex offender registries, and/or sex offender commitment. Findings in the curr ent study indicate that repeat victims are more likely to be sexually victimized at a younger age by familial perpetrators than women victimized by one perpetrator only Restorative justice could provide youth with the opportunity to share their stor y and validate that the offender is solely responsible for the violence, taking place in a less formal environment than standard prosecution processes (Daly & Stubbs, 2006). Nonetheless restorative justice tactics regarding sexual abuse are divisive, due to concerns regarding victim safety, offender exploitation of the process, victim inability to articulate their desired outcome, community norms discouraging offender assignment of blame, blended loyalties among friends and family, and the potential lack of impa ct on offender behavior (Daly & Stubbs,
124 2006). Research is in the preliminary stages assessing the effectiveness of restorative justice and gendered violence (Jli s ch, 2006), but emerging pilot programs should consider assessing the ir impact on likelihood of sexual repeat victimization In terms of a secondary program intervention on post -sexual victimization mental health existing programs such as sexual assault nurse examiner (SANE) programs and sexual assault response teams (SARTs) provide models for a ll medical and legal professionals on the front line, who are often the first sources of formal support victim s seek out (Campbell et al., 2009). Generally, survivors indicate experiences with mental health professionals impact their recovery more positive ly than interactions with the criminal justice system (Campbell et al., 2001; Ullman, 1996a, 1996b; Taylor & Harvey, 2010) Although victims experience secondary victimization in forensic or medical examinations, survivors are less likely to report feeling guilty, depressed, anxious, distrustful, and reluctan t to pursue further support than those disclosing to criminal justice support systems (Campbell, 2005, 2006; Campbell & Raja, 2005). In the present study, police engagement was protective against long t erm depression in sexual repeat victims, suggesting that access to formal justice systems may support post victimization recovery long -term Similarly, increasing funding and access to rape crisis centers would provide victims with improved support options and community advocacy (Campbell et al., 2009) Although therapeutic support options for victims are necessary, o ne study of women seeking assistance after domestic or sexual victimization found most reporting that pragmatic services, such as food, housi ng, and financial assistance were more helpful than professional counseling (Postmus, Severson, Berry, & Yoo, 2009). Advocates on the
125 front lines of service are aware of the diversity of survivor needs and can supplement the support friends, family, and th e police provide. Conclusion Sexual repeat violence and its impact on survivors are trenchant, complex problems that have resisted simple theoretical explanations or policy solutions. Application of general criminological theory and policy to interpersonal violence is challenging as evidence suggests that violent repeat victimization is unique. International crime surveys document that unlike other crime s sexual repeat victimization rates are fairly consistent across countries (Farrell & B ouloukos, 2001). In addition, sexual crimes have the highest rates of repeat victimization, with between 40% and 50% of total incidents representing repeat victimizations (Farrell & Bouloukos, 2001) Evaluations of victim safety and support policies for th is unique subtype of repeat victimization indicate lingering challenges One study examining respondents experiences with victimization, police, and other government agencies across 54 countries found r epeat victims were less likely to report to the polic e than other victims because they believed police were un willing or in capable of helping them (Van Dijk, 2001). Repeat victims were also more likely to report dissatisfaction with the police, particularly in industrialized countries. Those who did not report their victimization to the police indicated that they addressed the problem on their own accord perceived the issue to be inappropriate for the police, or reported fear or dislike of the police. On average, over 30% of victims of gender violence compla ined about the police being rude or incorrect (Van Dijk, 2001). The body of literature on sexual victimization extensively documents the likelihood of repeat victimization and the multiplicative psychological impact of repeat
126 sexual trauma. Extant research has contributed prospective examinations of risk and protective factors of repeat victimization among clinical and university populations, typically with separate emphases on psychological trauma, behaviora l, or situational variables. This current research complements these longitudinal assessments of clustered risk and protective factors through collectively examining a range of individual, situational, and socio cultural factors influencing repeat rape likelihood in a large, national rando m -digit -dialing sample. In addition, the present study extends prior research on post -sexual victimization mental health t hrough modeling individual, situational, and sociocultural predictors of depression among a sample of rape survivors, and then separat ely among women victimized by one offender versus those victimized by multiple distinct offenders. Findings indicate d that risk of repeat victimization by a new offender increases when victimization is perpetrated by a family member with greater physical force against younger individuals in a n ambiguous pre victimization context. Separate models of longterm depression risk for women victimized by one perpetrator only and repeat victims of rape by separate perpetrators detected unique risk and protective fa ctors. For survivors of rape victimization by one perpetrator long -term depression risk was predi cted by chronic health problems and taking time off after the victimization. Pregnancy as a result of the rape victimization, education, and income were protective against long term depression, and the use of physical violence as a coercion tactic in the primary victimization trended toward protecting against long term depression for survivors of rape by one perpetrator For repeat rape victims chronic he alth problems, being a woman of color, and quantity of informal support predicted greater depression
127 and police engagement and a n innocuous pre -victimization context shielded against long -term depression. Ineffective policy responses to sexual repeat victimization and recovery and the limited success in developing explanatory theories of sexual repeat victimization reflect our incomplete understanding of the origins of sexual repeat victimization. A rich body of literature examining risk and protective fac tors informed the current ecological assessment of models of sexual repeat victimization and long-term post -victimization depression. Mixed findings and methodological limitations impede conclusions on the appropriateness of ecological sexual repeat victimization and post -victimization mental health models. Nonetheless, the current study diversifies the evidence regarding vulnerabilities and protective factors surrounding sexual repeat victimization and trauma in a general population sample for continued theory development and policy application.
128 BIBLIOGRAPHY Abbey, A. (2002). Alcohol related sexual assault: A common problem among college students. Journal of Studies on Alcohol. Supplement, 14, 118 28. Acierno, R., Resnick, H. S., & Kilpatrick, D. G. (1997). Prevalence rates, case identification, and risk factors for sexual assault, physical assault, and domestic violence in men and women, part 1. Behavioral Medicine, 23(2), 53 66. Acierno, R., Resnic k, H. S., Kilpatrick, D. G., Saunders, B., & Best, C. L. (1999). Risk factors for rape, physical assault, and Posttraumatic Stress Disorder in women: Examination of differential multivariate relationships. Journal of Anxiety Disorders, 13(6), 541563. Ahr ens, C. (2006). Being silenced: The impact of negative social reactions on the disclosure of rape. American Journal of Community Psychology, 38, 263 274. Ahrens, C. E., Abeling, S., Ahmad, S., & Hinman, J. (2010). Spirituality and well being: The relation ship between religious coping and recovery from sexual assault. Journal of Interpersonal Violence, 25(7), 12421263. Aizer, A. (2010). The gender wage gap and domestic violence. The American Economic Review, 100(4), 18471859. Allison, P. D. (2009). Missing data. In R. E. Millsap & A. Maydeu Olivares (Eds.) The SAGE handbook of quantitative methods in psychology (pp. 72 90). Thousand Oaks, CA: SAGE Publications Inc. Amir, M. (1967). Victim precipitated forcible rape. Journal of Criminal Law, Criminol ogy and Police Science, 58(4), 493502. Andrews, B., Brewin, C. R., & Rose, S. (2003). Gender, social support, and PTSD in victims of violent crime. Journal of Traumatic Stress, 16(4), 421 427. Arata, C. M. (1999). Sexual repeat victimization and PTSD: A n exploratory study. Journal of Child Sexual Abuse, 8(1), 49 65. Arata, C. M. (2000). From child victim to adult victim: A model for predicting sexual revictimization. Child Maltreatment, 5 28 38. Arata, C. M. (2002). Child sexual abuse and sexual revi ctimization. Clinical Psychology: Science and Practice, 9(2), 135164. Atkeson, B. M., Calhoun, K. S., Resick, P. A., & Ellis, E. M. (1982). Victims of rape: Repeated assessment of depressive symptoms. Journal of Consulting and Clinical Psychology, 50, 96 102.
129 Banyard, V. L., Williams, L. M., & Siegel, J. A. (2001). The long term mental health consequences of child sexual abuse: An exploratory study of the impact of multiple traumas in a sample of women. Journal of Traumatic Stress 14, 697715. Barnes, J. E., Noll, J. G., Putnam, F. W., & Trickett, P. K. (2009). Sexual and physical repeat victimization among victims of severe childhood sexual abuse. Child Abuse & Neglect, 33(7), 412 420. Beccaria, C. (1963). Essay on crimes and punishment (H. Paolucci Trans.). Indianapolis: Bobbs Merrill. (Original work published 1764). Belsky, J. (1980). Child maltreatment: An ecological integration. American Psychologist, 35, 320335. Ben David, S. (2000). Needed: Victims victimology. In P. C. Friday & G. F. Kirchhoff (Eds.), Victimology at the transition from the 20th to the 21st century (pp. 55 72). Monchengladbach, Germany: Shaker Verlag & World Society of Victimology Publications. Bjrklund, K., Hkknen Nyholm, H., Huttunen, T., & Kunttu, K. (2010). Violence victimization among Finnish university students: Prevalence, symptoms and healthcare usage. Social Science and Medicine, 70, 1416 1422. Bograd, M. (1999). Strengthening domestic violence theories: Intersections of race, class, sexual orie ntation, and gender. Journal of Marital and Family Therapy, 25(3), 275 289. Borja, S. E., Callahan, J. L., & Long, P. J. (2006). Positive and negative adjustment and social support of sexual assault survivors. Journal of Traumatic Stress, 19, 905914. Bo werman, B. L., & OConnell, R. T. (1990). Linear statistical models: An applied approach (2nd ed.). Belmont, CA: Duxbury. Bownes, I. T., OGorman, E. C., & Sayers, A. (1991). Assault characteristics and posttraumatic stress disorder in rape victims. Acta Psychiatrica Scandinavia, 83, 27 30. Boyer, D., & Fine, D. (1992). Sexual abuse as a factor in adolescent pregnancy and child maltreatment. Child Abuse and Neglect, 19(12), 14011421. Brand, B. L., & Alexander, P. C. (2003). Coping with incest: The relat ionship between recollections of childhood coping and adult functioning in female survivors of incest. Journal of Traumatic Stress, 16 (3), 285293. Breitenbecher, K. (2001). Sexual repeat victimization among women. A review of the literature focusing on e mpirical investigations. Aggression & Violent Behavior, 6, 415 432.
130 Brener, N. D., McMahon, P. M., Warren, C. W., & Douglas, K. A. (1999). Forced sexual intercourse and associated health risk behaviors among female college students in the United States. J ournal of Consulting and Clinical Psychology, 67, 252259. Briere, J. (1995). Trauma symptom inventory manual Odessa, FL: Psychological Assessment Resources. Briere, J., & Jordan, C. E. (2009). Childhood maltreatment, intervening variables, and adult ps ychological difficulties in women: An overview. Trauma Violence Abuse, 10(4), 37588. Brofenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, Massachusetts: Harvard University Press. Bronfenbrenner U. (1977). Toward an experimental ecology of human development. American Psychologist, 32, 513 531. Brown, A. L., Testa, M., & Messman -Moore, T. L. (2009). Psychological consequences of sexual victimization resulting from force, incapacitation, and verb al coercion. Violence Against Women, 15, 898919. Brownmiller, S. (1975). Against our will: Men, women and rape New York: Simon and Schuster Inc.. Bryant Davis, T., Ullman, S. E., Tsong, Y., & Gobin, R. (2011). Surviving the storm: The role of social su pport and religious coping in sexual assault recovery of African American women. Violence Against Women, 17(12), 16011618. Bryant Davis, T., Ullman, S. E., Tsong, Y., Tillman, S., & Smith, K. (2010). Struggling to survive: Sexual assault, poverty, and me ntal health outcomes of African American women. American Journal of Orthopsychiatry, 80(1), 6170. Burge, S. K. (1988). Post traumatic stress disorder in victims of rape. Journal of Traumatic Stress, 1, 193-210. Burnam, A. M., Stein, J. A., Golding, J. M., Siegel, J. M., Sorensen, S. B., Forsythe, A. B. & Telles, C. A. (1988). Sexual assault and mental disorders in a community population. Journal of Consulting & Clinical Psychology 56, 843850. Burt, M. R., & Katz, B. L. K. (1987). Dimensions of recove ry from rape: Focus on growth outcomes Journal of Interpersonal Violence, 2, 57 81. Busch -Armendariz, N. B., Bell, H., DiNitto, D. M., & Neff, J. (2003, August). The Health Survey of Texans: A focus on sexual assault. Austin: University of Texas at Austi n, Institute on Domestic Violence and Sexual Assault.
131 Cahalan, D., Cisin, I. H., & Crossley, H. M. (1969). American drinking practices: A national study of drinking behavior and attitudes New Brunswick, NJ: Rutgers Center of Alcohol Studies. Campbell, R (2005). What really happened? A validation study of rape survivors help-seeking experiences with the legal and medical systems. Violence & Victims, 20 55 -68. Campbell, R. (2008). The psychological impact of rape victims experiences with the legal, me dical, and mental health systems. American Psychologist, 63, 702 717. Campbell, R., & Raja, S. (2005). The sexual assault and secondary victimization of female veterans: Helpseeking experiences in military and civilian social systems. Psychology of Women Quarterly, 29 97 106. Campbell, R., Ahrens, C., Sefl, T., Wasco, S. M., & Barnes, H. E. (2001). Social reactions to rape victims: Healing and hurtful effects on psychological and physical health outcomes. Violence & Victims, 16 287302. Campbell, R., Dworkin, E., & Cabral, G. T. (2009). An ecological model of the impact of sexual assault on womens mental health. Journal of Interpersonal Violence, 10(3), 22546. Campbell, R., Sefl, T., Barnes, H. E., Ahrens, C. E., Wasco, S. M., & Zaragoza Diesfeld, Y (1999). Community services for rape survivors: Enhancing psychological well -being or increasing trauma? Journal of Consulting and Clinical Psychology, 67, 847858. Carnevale, A. P., Rose, S. J., & Cheah, B. (2011). The college payoff: Education, occupat ions, lifetime earnings Washington DC: The Georgetown University Center on Education and the Workforce. Cartwright, D., & Zander, A. (1968). Group dynamics. New York: Harper and Row. Casey, E. A., & Nurius, P. S. (2005). Trauma exposure and sexual revic timization risk: Comparisons across single, multiple incident, and multiple perpetrator victimizations. Violence Against Women, 11, 505 530. Central Intelligence Agency (2013). Country comparison: Distribution of family income Gini Index. The World Factbook. Washington DC: Author. Retrieved from: https://www.cia.gov/library/publications/the -world -factbook/rankorder/2172rank.html Champion, H. L., Foley, K. L ., Durant, R. H., Hensberry, R., Altman, D., & Wolfson, M. (2004). Adolescent sexual victimization, use of alcohol and other substances, and other health risk behaviors. Journal of Adolescent Health, 35(4), 321328. Chan, K. (2011). Gender differences in self -reports of intimate partner violence: A review. Aggression and Violent Behavior, 16, 167 175.
132 Charuvastra, A., & Cloitre, M. (2008). Social bonds and posttraumatic stress disorder. Annual Review of Psychology, 59 301328. Classen, C., Field, N. P., Koopman, C., Nevill Manning, K., & Spiegel, D. (2001). Interpersonal problems and their relationship to sexual revictimization among women sexually abused in childhood. Journal of Interpersonal Violence, 16, 495509. Cloitre, M., Cohen, L. R., & Scarvalone, P. (2002). Understanding revictimization among childhood sexual abuse survivors: An interpersonal schema approach. Journal of Cognitive Psychotherapy, 16(1), 91 112. Cloutier, S., Martin, S. L., & Poole, C. (2002). Sexual assault among North Carolina women: prevalence and health risk factors. Journal of Epidemiology and Community Health, 56, 265271. Clum, G. A., Nishith, P., & Calhoun, K. S. (2002). A preliminary investigation of alcohol use during trauma and peritraumatic reactions in female sexual assault victims. Journal of Traumatic Stress, 15, 321328. Cohen, L. & Felson, J. (1979). Social change and crime rates trends: A routine activity approach. American Sociological Review, 44, 588608. Coid, J., Pet ruckevitch, A., Feder, G., Chung, W., Richardson, J., & Moorey, S. (2001). Relation between childhood sexual and physical abuse and risk of revictimisation in women: A cross sectional survey. Lancet 358, 450454. Collins, M. E. (1998). Factors influencing sexual victimization and repeat victimization in a sample of adolescent mothers. Journal of Interpersonal Violence, 13(1), 3 24. Copeland, K. R. (1996). The effects of childhood abuse and family environment on attitudes, behaviors, and adult revictimization Unpublished doctoral dissertation. Los Angeles: University of Southern California. Crenshaw, K. (1991). Mapping the margins: Intersectionality, identity politics, and violence against women of color. Stanford Law Review, 43(6), 12411299. Cuevas, C A., Finkelhor, D., Clifford, C., Ormrod, R. K., & Turner, H. A. (2010) Psychological distress as a risk factor for re -victimization in children. Child Abuse & Neglect, 34, 235 243. Culbertson, K. A., & Dehle, C. (2001). Impact of sexual assault as a function of perpetrator type. Journal of Interpersonal Violence, 16, 9921007. Daigle, L. E., Fisher, B. S., & Cullen, F. T. (2008). The violent and sexual victimization of college women: Is repeat victimization a problem? Journal of Interpersonal Violence, 23, 12961313.
133 Daly, K. & Stubbs, J. (2006). Feminist engagement with restorative justice. Theoretical Criminology, 10(1), 9 28. Davis, R. C., Brickman, E., & Baker, T. (1991). Supportive and unsupportive responses of others to rape victims: effects on concurrent victim adjustment. American Journal of Community Psychology, 19, 443451. Day, K. (1994). Conceptualizing womens fear of sexual assault on campus: A review of causes and recommendations for change. Environment and Behavior, 26, 742765. Deitz, S. R., Littman, M., & Bentley, B. J. (1984). Attribution of responsibility for rape: The influence of observer empathy, victim resistance, and victim attractiveness. Sex Roles 10(3 4), 261280. DeMaris, A., & Kaukinen, C. (2005). Violent victimization a nd womens mental and physical health: Evidence from a national sample. Journal of Research in Crime & Delinquency, 42, 384411. DeNavas-Walt, C., Proctor, B. D., & Smith, J. C. (2009). Income, poverty, and health insurance coverage in the United States: 2009. Washington, DC: U.S. Government Printing Office. Desai, S., Arias, I., Thompson, M. P., & Basile, K. C. (2002). Childhood victimization and subsequent adult repeat victimization assessed in a nationally representative sample of women and men. Violen ce and Victims 17, 639 653. Dubin, J. A. & Rivers, D. (1989). Selection bias in logit and probit models. Sociological Methods and Research, 18(2/3), 360390. Dussich, J. P. (2003). History, overview and analysis of American victimology and victim servic es education. American Society of Victimology 4 18 25. Ellenberger, H. (1955). Psychological relationships between the criminal and his victim. Archives of Criminal Psychodynamics 2 257290. Elliott, D. M., Mok, D., & Briere, J. (2004). Adult sexual assault: Prevalence, symptomatology, and sex differences in the general population. Journal of Traumatic Stress 17, 203211. Ellis, E. M., Atkeson, B. M. & Calhoun, K. S. (1981). Assessment of long term reactions in victims of rape. Journal of Abnormal Psychology, 90, 263266. Ellis, E. M., Atkeson, B. M., & Calhoun, K. S. (1982). An examination of differences between multiple and single incident victims of sexual assault. Journal of Abnormal Psychology, 91, 221 224.
134 Fargo, J. D. (2009). Pathways to adult sexual revictimization: Direct and indirect behavioral risk factors across the lifespan. Journal of Interpersonal Violence, 24(11), 17711791. Farrell, G. & Bouloukos, A. C. (2001). International overview: A cross -national comparison of rates of repeat victimization. Crime Prevention Studies, 12, 5 25. Farrell, G. & Pease, K. (2001). Why repeat victimization matters In In Farrell, G. and Pease, K. (Eds.) Repeat Victimization. Monse y, NY: Criminal Justice Press. Farrigan, T. (2012). Rural poverty and well -being. Washington DC: USDA, Economic Research Service. Retrieved from: http://www.ers.usda.gov/topics/rural -economy-population/rural poverty-well -being.aspx#.UZk_WrWsiSo Fattah, E. (1991). Understanding criminal victimisation. Canada: Prentice Hall. Fattah, E. A. (1994). Some problematic concepts, unjustified cri ticism and popular misconceptions. In G. F. Kirchhoff, E. Kosovski, & H. J. Schneider (Eds.), International debates of victimology (pp. 82103). Moenchengladbach, Germany: World Society of Victimology. Fattah, E. A. (2000). Victimology: Past, present and future. Criminologie 33(1), 17 46. Filipas, H. H., & Ullman, S. E. (2006). Child sexual abuse, coping responses, self -blame, posttraumatic stress disorder, and adult sexual repeat victimization. Journal of Interpersonal Violence 21, 652 672. Finkelhor, D., & Browne, A. (1985). The traumatic impact of child sexual abuse: A conceptualization. American Journal of Orthopsychiatry, 55, 530541. Finkelhor, D., & Browne, A. (1988). Child sexual abuse: A review and conceptualization. In G. T. Hotaling, D. Fink elhor, J. T. Kirkpatrick, & M. A. Straus (Eds.), Family abuse and its consequences: New directions in research (pp. 270284). Newbury Park, CA: Sage. Finkelhor, D., Ormrod, R. K., & Turner, H. A. (2007). Poly -victimization: A neglected component in child victimization trauma. Child Abuse & Neglect, 31, 7 26. Fisher, B. S., Cullen, F. T., & Turner, M. G. (2000). The sexual victimization of college women. Washington, DC: U.S. Department of Justice, Office of Justice Programs. Fisher, B. S., Daigle, L. E., & Cullen, F. T. (2008). Rape against women: What can research offer to guide the development of prevention programs and risk reduction interventions? Journal of Contemporary Criminal Justice, 24, 163177. Fisher, B. S., Daigle, L. E., & Cullen, F. T. (2009). What distinguishes single from recurrent sexual victims? The role of lifestyle routine activities and first incident characteristics. Justice Quarterly, 27 (1), 102129.
135 Fisher, B., & Cullen, F. (2000). Measuring the sexual victimization of women: Evolution, current controversies, and future research. In Measurement and analysis of crime and justice, criminal justice 2000 (Volume 4, pp. 317390). Washington, DC: U.S. Department of Justice, National Institute of Justice. Fontana, V. O. (1973). Somew here a child is crying New York: Macmillan Publishing Company. Fortier, M. A., DiLillo, D., Messman -Moore, T. L., Peugh, J., DeNardi, K. A., & Gaffey, K. J. (2009). Child sexual abuse and revictimization: The mediating role of coping and trauma symptoms. Psychology of Women Quarterly, 33, 308 320. Frank, E., & Stewart, B. D. (1984). Depressive symptoms in rape victims: A revisit. Journal of Affective Disorders, 7 77 85. Frank, E., Turner, S. M., & Stewart, B. D. (1980). Initial response to rape: The impact of factors within the rape situation. Journal of Behavioral Assessment, 2, 39 53. Frazier, P. A., Mortenson, H., & Steward, J. (2005). Coping strategies as mediators of the relations among perceived control and distre ss in sexual assault survivors. Journal of Counseling Psychology 52, 267 278. Frazier, P., Tashiro, T., Berman, M., Steger, M., & Long, J. (2004). Correlates of levels and patterns of positive life changes following sexual assault. Journal of Consulting and Clinical Psychology, 72, 19 30. Friesen, M. D., Woodward, L. J., Horwood, L. J., & Fergusson, D. M. (2009). Childhood exposure to sexual abuse and partnership outcomes at age 30. Psychological Medicine, 40, 679688. Frieze, I. H., & Browne, A. (1989) Violence in marriage. In L. Ohlin & M. Tonry (Eds.), Family violence (pp. 163 218). Chicago: University of Chicago Press. Frohmann, L. (1991). Discrediting victims allegations of sexual assault: Prosecutorial accounts of case rejections. Social Problem s, 38, 213226. Gabor, T., & Mata, F. (2004). Victimization and repeat victimization over the life span: A predictive study and implications for policy. International Review of Victimology, 10 (3), 193221. Garofalo, B. R. (1914). Criminology. Boston: Little, Brown & Company. George, W. H., & Martnez, L. J. (2002). Victim blaming in rape: Effects of victim and perpetrator race, type of rape, and participant racism. Psychology of Women Quarterly, 26, 110119. Gidycz, C. A., Coble, C. N., Latham, L. & Layman, M. J. (1993). Sexual assault experiences in adulthood and prior victimization experiences: A prospective analysis. Psychology of Women Quarterly, 17, 151 168.
136 Gidycz, C. A., Hanson, K., & Layman, M. J. (1995). A prospective analysis of the relationships among sexual assault experiences. Psychology of Women Quarterly, 19, 5 29. Goldman, D., G. Oroszi & F. Ducci. 2005. The genetics of addictions: uncovering the genes. Nature Reviews Genetics, 6 521 532. Goodman, L. A., Koss, M. P., & Russo, N. F. (1993). Violence against women: Physical and mental health effects. Part I: Research findings. Applied and Preventative Psychology, 2, 79 89. Gottfredson, M. (1984). Victims of crime: The dimensions of risk. Home office research study no. 81. London: H.M. Stationery Office. Graham, J. W. (2009). Missing data analysis: Making it work in the real world. Annual Review of Psychology, 60, 549576. Graham, J. W. (2012). Missing data: Analysis and design. New York: Springer. Graham, J. W., Olchowski, A. E., & Gilreath, T. D. (2007). How many imputations are really needed? Some practical clarifications of multiple imputation theory. Prevention Science, 8 206 213. Grauerholz, L. (2000). An ecological approach to understanding sexual repeat victimization: Linkin g personal, interpersonal, and sociocultural factors and processes. Child Maltreatment, 5, 5 17. Greene D. M., & Navarro, R. L. (1998). Situation -specific assertiveness in the epidemiology of se xual victimization among university women. Psychology of Women Quarterly, 22, 589 604. Griffin, M. J., & Read, J. P. (2012). Prospective effects of method of coercion in sexual victimization across the first college year. Journal of Interpersonal Violence 27(12), 25032524. Gutner, C. A., Rizvi, S. L., Monson, C. M., & Resick, P. A. (2006). Changes in coping strategies, relationship to perpetrator, and posttraumatic distress in female crime victims. Journal of Traumatic Stress, 19, 813823. Harrell, F. E. (1984). Regression modeling strategies for improved prognostic prediction. Statistics in Medicine, 3 143152. Harvey, M. R. (1996). An ecological view of psychological trauma and trauma recovery. Journal of Traumatic Stress, 9 3 23. Heck man, J. J. (1974). Shadow prices, market wages, and labor supply. Econometrica, 42, 679694.
137 Heise, L. (1998). Violence against women: An integrated, ecological framework. Violence Against Women, 4, 262290. Herman, J. L. (1992). Complex PTSD: A syndrome in survivors of prolonged and repeated trauma. Journal of Traumatic Stress, 5, 377391. Higgins, E. T. (1989). Knowledge accessibility and activation: Subjectivity and suffering from unconscious sources. In J. S. Ulleman & J. A. Bargh (Eds.) Unintended T hought (pp. 75123). New York: Guilford Press. Himelein, M. J. (1995). Risk factors for sexual victimization in dating: A longitudinal study of college women. Psychology of Women Quarterly, 19, 3143. Hindelang, M. J., Gottfredson, M. R., & Garofalo, J. ( 1978). Victims of personal Crime: An empirical foundation for a theory of personal victimization. Cambridge, MA: Ballinger. Hosser, D., Raddatz, S., & Windzio, M. (2007). Child maltreatment, repeat victimization, and violent behavior. Violence and Victims 22, 318 333. Jacques Tiura A. J., Tkatch R., Abbey A., & Wegner R. (2010). Disclosure of sexual assault: Characteristics and implications for posttraumatic stress symptoms among African American and Caucasian survivors. Journal of Trauma & Dissociation, 11(2), 174 192. Janus, E. (2006). Failure to protect: Americas sexual predator laws and the rise of the preventive state Ithaca, New York: Cornell University Press. Johnson, D. M., Shehan, T. C., & Chard, K. M. (2003). Personality disorders, coping st rategies, and posttraumatic stress disorder in women with histories of childhood sexual abuse. Journal of Child Sexual Abuse, 12(2), 19 39. Johnson, N. L. & Johnson, D. M. (in press). Factors influencing the relationship between sexual trauma and risky sexual behavior in college students. Journal of Interpersonal Violence. Jones, J. S., Alexander, C., Wynn, B. N., Rossman, L., & Dunnuck, C. (2009). Why women dont report sexual assault to the police: The influence of psychosocial variables and traumati c injury. The Journal of Emergency Medicine, 36(4), 417 424. Jordan, C. E. (2004). Intimate partner violence and the justice system: An examination of the interface. Journal of Interpersonal Violence, 19, Jlisch, S. (2006). Views of justice among survivors of historical child abuse: Implications for restorative justice in New Zealand. Theoretical Criminology, 10(1), 125138. Karmen, A. (2012). Crime victims: An introduction to victimology Belmont, CA: Wadsworth publishing company.
138 Katz, J., May, P., Srensen, S., & DelTosta, J. (2010). Sexual revictimization during womens first year of college: Self -blame and sexual refusal assertiveness as possible mechanisms. Journal of Interpersonal Violence, 25(11), 21132126. Kayse n, D., Neighbors, C., Martell, J., Fossos, N., & Larimer, M. E. (2006). Incapacitated rape and alcohol use: A prospective analysis. Addictive Behaviors 31(10), 18201832. Kearns, M. C., & Calhoun, K. S. (2010). Sexual revictimization and interpersonal ef fectiveness. Violence and Victims, 25 504 517. Kellogg, N. D. & Hoffman, T. J. (1997). Child sexual revictimization by multiple perpetrators. Child Abuse and Neglect, 21, 953 964. Kilpatrick, D. G. (2004). What is violence against women: Defining and me asuring the problem. Journal of Interpersonal Violence 19, 12091234. Kilpatrick, D. G., Amstadter, A. B., Resnick, H. S., & Ruggiero, K. J. (2007). Rape related PTSD: Issues and interventions. Psychiatric Times, 24 50 58. Kilpatrick D. G., Resnick, H. S., Saunders, B. E., Best, C. L. (1998). Rape, other violence against women,and posttraumatic stress disorder. In B. P. Dohrenwend (Eds.), Adversity, Stress, and Psychopathology (pp. 161 176). London: Oxford University Press. Kilpatrick, D. G., Veronen, L. J., & Best, C. L. (1984). Factors predicting psychological distress among rape victims. In C. R. Figley (Ed.), Trauma and its wake (pp. 113141). Bristol, PA: Brunner/Mazel. Kimerling, R., Alvarez, J., Pavao, J., Kaminski, A., Baumrind, N. (2007), Epidemiology and consequences of womens repeat victimization. Womens Health Issues, 17 101 6. Kirchhoff, G. (2003). A history of victimology. Presentation at the Post -graduate Course on Victimology, Victim Assistance and Criminal Justice, Stellenbosch University, South Africa, 7 11 July. Knowles, K. B. (1993). Sexual revictimization: a comparison of single -incident versus recidivist rape victims on symptom severity, definitions, attributions, coping strategies, and precautionary behaviors. Unpublished doctoral dissertation. Auburn, AL: Auburn University. Koss, M. P. Abbey, A., Campbell, R., Cook, S., Norris, J., Testa, M., Ullman, S., West, C., & White, J. (2006). The Sexual Experiences Short Form Victimization (SES-SFV). Tucson, AZ: University of Ari zona.
139 Koss, M. P., Gidycz, C. A., & Wisniewski, N. (1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Psychology, 55, 162170. K rah, B., Scheinberger Olwig, Waizenheffer, E., & Kolpin, S. (1999). Childhood sexual abuse and revictimization and adolescence. Child Abuse and Neglect, 23, 383 394. Krebs, C. P., Lindquist, C. H., Warner,T. D., Fisher, B. S., & Martin,S L. (2009). The differential risk factors of physically forced and alcohol -or other drug -enabled sexual assault among university women. Violence and Victims, 24(3), 302322. Leonard, W. (1992). Incest and revictimization: the relationship between aspects of the incest experience, self -efficacy, interpersonal dependency, interpersonal trust, sexual self -esteem, sexual behavior, and revictimization. Unpublished doctoral dissertation. Philadelphia, PA: Temple University. Lewis, R. (2004). Making justice wor k: Effective legal interventions for domestic violence. British Journal of Criminology, 44, 204224. Littleton, H. L., & Henderson, C. E. H. (2009). If she is not a victim, does that mean she was not traumatized? Evaluation of predictors of PTSD symptomat ology among college rape victims. Violence Against Women, 15, 148 167. Littleton, H. L., Grills Taquechel, A. E., Buck, K. S., Rosman, L., & Dodd, J. C. (in press). Health risk behavior and sexual assault among ethnically diverse women. Psychology of Wome n Quarterly Logan, C., Holcombe, E., Ryan, S., Manlove, J., & Moore, K. (2007). Childhood sexual abuse and pregnancy: A white paper Washington DC: Child Trends Inc. Lombardi, K. (2009, December 1). Sexual assault on campus shrouded in secrecy. The Center for Public Integrity. Retrieved from http://www.publicintegrity.org/investigations/campus_assault/articles/entry/1838/ Macy, R. J. (2008). A research agenda for sexual repeat victimization: Priority areas and innovative statistical methods. Violence Against Women, 14(10), 11281147. Macy, R. J., Nurius, P., & Norris, J. (2006). Responding in their best interests: Contextualizing womens coping with acquaintance sexual aggression. Violence Against Women, 12, 478500. Maker, A. H., Kemmelmeier, M., & Peterson, C. (2001). Child sexual abuse, peer sexual abuse, and sexual assault in adulthood: A multi -risk model of revictimization. Journal of Traumatic Str ess, 14, 351368.
140 Mandoki, C. A. & Burkhart, B. R. (1989). Sexual victimization: is there a vicious cycle? Violence and Victims, 4 179 190. Marhoefer Dvorak, S., Resick, P. A., Hutter, C. K., & Girelli, S. A. (1988). Single versus multiple incident rap e victims: a comparison of psychological reactions to rape. Journal of Interpersonal Violence 3, 145 160. Martin, D. (1978). Battered women: Society's problem. In J. R. Chapman & M. Gates (Eds.), The victimization of women (pp. 111141). Beverly Hills, CA: Sage. Marx, B. P., Calhoun, K. S., Wilson, A. E., & Meyerson, L. A. (2001). Sexual revictimization prevention: An outcome evaluation. Journal of Consulting and Clinical Psychology, 69(1), 2532. Marx, B. P., Heidt, J. M ., & Gold, S. D. (2005). Perceived uncontrollability and unpredictability, self regulation, and sexual repeat victimization. Review of General Psychology, 9, 67 90. Masho, S. W., & Ahmed, G. (2007). Age at sexual assault and posttraumatic stress disorder among women: Prevalence, correlates, and implications for prevention. Journal of Women's Health 16(2), 262271. Mawby, R. B. & Walklate, S. (1994). Critical victimology. London: Sage Publications. Mayall, A., & Gold, S. R. (1995). Definitional issues an d mediating variables in the sexual revictimization of women sexually abused as children. Journal of Interpersonal Violence, 10(1), 26 42. McCutcheon, V. V., Sartor, C. E., Pommer, N. E., Bucholz, K. K., Nelson, E. C., Madden, P. A. F., & Heath, A. C. (2010). Age at trauma exposure and PTSD risk in young adult women. Journal of Traumatic Stress, 23(6), 811 814. McFarlane, J., Malecha, A., Watson, K., Gist, J., Batten, E., Hall, I., & Smith, S. (2005). Intimate partner sexual assault again st women: Frequency, health consequences, and treatment outcomes. Obstetrics and Gynecology, 105, 99 108. Mears, D. P., Carlson, M. J., Holden, G. W., & Harris, S. D. (2001). Reducing domestic violence repeat victimization: The effects of individual and c ontextual factors and type of legal intervention. Journal of Interpersonal Violence, 16 Mendelsohn, B. (1940). Rape in criminology. Translated and cited in The victim and his criminal (S. Schafer, 1968). New York: Random House. Mendelsohn, B. (1963). The origin of the doctrine of victimology. Excerpta Criminal, 3, 239244.
141 Merrill, L. L., Newell, C. E., Thomsen, C. J., Gold, S. R., Milner, J. S., Koss, M. P., & Rosswork, S. G. (1999). Childhood abuse and sexual revictimization in a female Navy recruit sample. Journal of Traumatic Stress, 12, 211 225. Merrill, L. L., Thomsen, C. J., Sinclair, B. B, Gold, S. R., & Milner, J. S. (2001). Predicting the impact of child sexual abuse on women: The role of abuse severity, parental support, and coping strategies. Journal of Consulting and Clinical Psychology, 69, 9921006. Messman Moore, T. L., & Brown, A. L. (2006). Risk perception, rape and sexual revictimization: A prospective study of college women. Psychology of Women Quarterly, 30, 159 172. M essman Moore, T. L., & Long, P. J. (2000). Child sexual abuse and repeat victimization in the form of adult sexual abuse, adult physical abuse, and adult psychological maltreatment. Journal of Interpersonal Violence 15, 489502. Messman Moore, T. L., & L ong, P. J. (2003). The role of childhood sexual abuse sequelae in the sexual revictimization of women: An empirical review and theoretical reformulation. Clinical Psychology Review, 23 537571. Messman Moore, T. L., Long, P. J., & Siegfried, N. J. (2000). The revictimization of child sexual abuse survivors: An examination of the adjustment of college women with child sexual abuse, adult sexual assault, and adult physical abuse. Child Maltreatment, 5 1827. Messman Moore, T. L., Ward, R. M., & Brow n, A. L. (2009). Substance use and PTSD symptoms impact the likelihood of rape and revictimization in college women. Journal of Interpersonal Violence, 24, 499 521. Meyer, C. B., & Taylor, S. E. (1986). Adjustment to rape. Journal of Personality and Social Psychology, 50, 1226 1234. Miller, A. K., Handley, I. M., Markman, K. D, & Miller, J. H. (2010). Deconstructing self -blame following sexual assault: The critical roles of cognitive content and process. Violence Against Wo men, 16, 11201137. Miller, J., Moeller, D., Kaufman, A., DiVasto, P., Pathak, D. & Christy, J. (1978). Recidivism among sexual assault victims. American Journal of Psychiatry, 135, 1103 1104. Mller, A. S., Bckstrm, T., Sngergaard, H. P., & Helstrm, L. (in press). Patterns of injury and reported violence depending on relationship to assailant in female Swedish sexual assault victims. Journal of Interpersonal Violence. Murphy, S. B., Potter, S. J., Pierce -Weeks, J., Stapleton, J. G. & Wiesen Martin, D. (2011). An examination of SANE data: Clinical considerations based on victim assailant relationship. Journal of Forensic Nursing, 7, 137144.
142 Nagel, W. H. (1963). The notion of victimology in criminology. Excerpta Criminologica, 3, 245247. Najdowski, C. J. & Ullman, S. E. (2011) The effects of revictimization on coping and depression in female sexual assault victims. Journal of Traumatic Stress, 24(2), 218 221. Nes, R. B., Czajkowski, N. O., Rysamb, E., rstavik. R. E., Tambs, K., Reichborn -Kjenneru d, T. (2012). Major depression and life satisfaction: A populationbased twin study. Journal of Affective Disorders, 144 (1/2), 5158. Neville, H. A., Heppner, M. J., Oh, E., Spanierman, L. B., & Clark, M. (2004). General and culturally specific factors in fluencing Black and White rape survivors self -esteem. Psychology of Women Quarterly, 28, 83 94. Noll, J. G., Horowitz, L. A., Bonanno, G. A., Trickett, P. K., & Putnam, F. W. (2003). Revictimization and self -harm in females who experienced childhood sexu al abuse. Journal of Interpersonal Violence 18(12), 1452 1471. Norris, F. H. (1992). Epidemiology of trauma: Frequency and impact of different potentially traumatic events on different demographic events. Journal of Consulting and Clinical Psychology, 60, 409418. Nurius, P. S., Norris, J., Young, D. S., Graham, T. L., & Gaylord, J. (2000). Interpreting and defensively responding to threats: Examining appraisals and coping with acquaintance sexual aggression. Violence and Victims, 15 187208. O'Connell, M. (2008). Victimology: A Social Science in Waiting?. International Review of Victimology 15(2), 91 104. Orchowski, L. M., Untied, A. S., & Gidycz, C. A. (in press). Social reactions to disclosure of sexual victimization and adjustment among s urvivors of sexual assault. Journal of Interpersonal Violence Paolucci, E. O., Genuis, M. L., & Violato, C. (2001). A meta analysis of the published research on the effects of child sexual abuse. The Journal of psychology 135 (1), 17 36. Peterson, C., & Seligman, M. E. (1983). Learned helplessness and victimization. Journal of Social Issues, 39(2), 103116. Pezdek, K., & Taylor, J. (2002). Memory for traumatic events. In M. L. Eisen, G. S. Goodman, & J. A. Quas (Eds.), Memory and suggestibility in the forensic interview (pp. 165184). Mahwah, NJ: Lawrence Erlbaum and Associates. Planty, M., Langton, L., Krebs, C., Berzofsky, M. & Smiley McDonald, H. (2013). Female victims of sexual violence, 1994-2010. Washington DC: Bureau of Justice Statistics. Retri eved from: http://www.bjs.gov/content/pub/pdf/fvsv9410.pdf
143 Postmus, J. L., Severson, M., Berry, M., & Yoo, J. A. (2009). Women's experiences of violence and seeking help. Violence Against Wom en, 15(7), 85268. Proulx, J., Koverola, C., Fedorowicz, A. & Kral, M. (1995). Coping strategies as predictors of distress in survivors of single and multiple sexual victimization and nonvictimized controls. Journal of Applied Social Psychology, 25, 1464 1483. Reckless, W. (1967). The crime problem. New York: Appleton -Century Drafts. Reese Weber, M., & Smith, D. M. (2011). Outcomes of child sexual abuse as predictors of later sexual victimization. Journal of Interpersonal Violence, 26(9), 18841905. Res nick, H. S., Kilpatrick, D. G., Dansky, B. S., Saunders, B. E., & Best, C. L. (1993). Prevalence of civilian trauma and posttraumatic stress disorder in a representative national sample of women. Journal of Consulting and Clinical Psychology 61(6), 984991. Resnick, H. S., Walsh, K., Schumacher, J. A., Kilpatrick, D. G., & Acierno, R. (2013). Prior substance abuse and related treatment history reported by recent victims of sexual assault. Addictive Behaviors, 38 (4), 20749. Resnick, H. S., Yehuda, R., & Acierno, R. (1997). Acute post rape plasma cortisol, alcohol use, and PTSD symptom profile among recent rape victims. Annals New York Academy of Sciences, 821, 433436. Riggs, D. S., Kilpatrick, D. G., & Resnick, H. S. (1992). Long -term psychological dist ress associated with marital rape and aggravated assault: A comparison to other crime victims. Journal of Family Violence, 7 (4), 283296. Rock, P. (2002). On becoming a victim. In C. Hoyle & R. Young (Eds.) New visions of crime victims (pp. 1 22). Oxford, UK: Hart. Rock, P. (2007). Theoretical perspectives on victimisation. In S. Walklate (Ed.) Handbook of Victims and Victimology (pp. 37 61). New York: Willan Publishing. Roodman, A. A., & Clum, G. A. (2001). Revictimization rates and method variance: A m eta analysis. Clinical Psychology Review 21, 183204. Rubin, D.B. (1987). Multiple imputation for nonresponse in surveys New York: Wiley. Runtz, M., & Schallow, J. (1997). Social support and coping strategies as mediators of adult adjustment following childhood maltreatment. Child Abuse and Neglect, 21, 211 226. Russell, D. (1986). The secret trauma: Incest in the lives of girls and women. New York: Basic Books.
144 Samuels Dennis, J. A., Ford Gilboe, M., Wilk, P., Avison, W. R., & Ray, S. (2010). Cumulat ive trauma, personal and social resources, and post -traumatic stress symptoms among income assisted single mothers. Journal of Family Violence, 25, 603617. Sandberg, D. A., Matorin, A. I. & Lynn, S. J. (1999). Dissociation, posttraumatic symptomatology, and sexual revictimization: a prospective examination of mediator and moderator effects. Journal of Traumatic Stress, 12, 127 138. Sanders, B., & Moore, D. L. (1999). Childhood maltreatment and date rape. Journal of Interpersonal Violence, 14, 115124. Sarason, I. G., Sarason, B. R., Shearin, E. N., & Pierce, G. R. (1987). A brief measure of social support: Practical and theoretical implications. Journal of Social and Personal Relationships 4 (4), 497510. Schaaf, K. K., & McCanne, T. R. (1998). Relationship of childhood sexual, physical, and combined sexual and physical abuse to adult victimization and posttraumatic stress disorder. Child Abuse and Neglect, 22(11), 1119 1133. Schafer, S. (1968). The victim and his criminal: A study in functional responsibility New York: Random House. Schneider, H. J. (1992). Life in societal no -mans land: Aboriginal crime in Central Australia. International Journal of Offender Therapy and Comparative Criminology 36 5 19. Schneider, H. J. (2001). Victimologi cal developments in the world during the past three decades (I): A study of comparative victimology. International Journal of Offender Therapy and Comparative Criminology 45(4), 449 468. Sciolla, A., Glover, D. A., Loeb, T. B., Zhang, M., Myers, H. F., & Wyatt, G. E. (2011). Childhood sexual abuse severity and disclosure as predictors of depression among adult African American and Latina women. The Journal of Nervous and Mental Disease, 199(7), 471477. Sedgwick, J. L. (2006). Criminal victimization in the United States, 2005 statistical tables: National crime victimization survey Washington, DC: US Department of Justice, Bureau of Justice Statistics. Retrieved from: http://www.ojp.usdoj.gov/bjs/pub/pdf/cvus05.pdf Shaefer, H. L., & Edin, K. (in press). Rising extreme poverty in the United States and the response of federal means tested transfer programs. Social Service Review Smith, G. C., Kohn, S. J., Savage -Stevens, S. E., Finch, J. J., Ingate, R., & Lim, Y. O. (2000). The effects of interpersonal and personal agency on perceived control and psychological wellbeing in adulthood. The Gerontologist, 40(4), 458 468. Smits, J. (2003). Estimating the Heckman two -step procedure to control for selection bias with SPSS. Retrieved from : http://home.planet.nl/~smit9354/selbias/heckman.txt
145 Sokoloff, N. J., & Dupont, I. (2005). Domestic Violence at the Intersections of Race, Class, and Gender Challenges and Contributions to Understanding Violence Against Marginalized Women in Diverse Communities. Violence against Women, 11(1), 38 64. Sorenson, S. B., & Siegel, J. M. (1992). Gender, ethnicity, and sexual assault: Findings from a Los Angeles study. Journal of Social Issues, 48, 93104. Sparkes, R. F. (1982). Research on victims of crime: Accomplishments, issues and new directions. Crime and Delinquency Issues: Monograph Series, US Department of Health & Human Services, Maryland, USA. Steel, J., Sanna, L., Hammond, B., Whipple, J., & Cross, H. (2004). Psychological sequelae of childhood sexual abuse: abuse -related characteristics, coping strategies, and attributional style. Child Abuse & Neglect, 28, 785 801. Stermac, L., Del Bove, G., & Addison, M. (2001). Violence, injury, and presentation patterns in spousal sexual assaults. Violence Against Women 7 (11), 12181233. Stone, V. E., Cosmides, L., Tooby, J., Kroll, N.,& Knight, R. (2002). Selective impairment of reasoning about socia l exchange in a patient with bilateral limbic system damage. PNAS 99, 1153111536. Straus, M. A., Hamby, S. L., Boney McCoy, S., & Sugarman, D. B. (1996). The revised conflict tactics scales (CTS2): Development and preliminary psychometric data. Journal of Family Issues, 17 (3), 283316. Taylor, J. E., & Harvey, S. T. (2010). A meta analysis of the effects of psychotherapy with adults sexually abused in childhood. Clinical Psychology Review 30(6), 749767. Temple, J. R., Weston, R., Rodriguez, B. F., & Marshall, L. L. (2007). Differing effects of partner and nonpartner sexual assault on women's mental health. Violence Against Women, 13(3), 285297. Testa, M., & Dermen, K. H. (1999). The differential correlates of sexual coercion and rape. Journal of I nterpersonal Violence, 14, 548561. Testa, M., Hoffman, J. H., & Livingston, J. A. (2010). Alcohol and sexual risk behaviors as mediators of the sexual victimization revictimization relationship. Journal of Consulting and Clinical Psychology 78(2), 249. Tjaden, P. & Thoennes, N. (1998). Prevalence, incidence, and consequences of violence against women: Findings from the national violence against women survey (NCJ 172837). Washington, DC: United States Department of Justice, National Institute of Justice.
146 Tucker, J. S., Wenzel, S. L., Straus, J. B., Ryan, G. W., & Golinelli, D. (2005). Experiencing interpersonal violence perspectives of sexually active, substanceusing women living in shelters and low income housing. Violence Against Women 11(10), 13191 340. U.S. Census Bureau (1992a). We asked... you told us: Hispanic origin. Washington, DC: Author. Retrieved from: http://www.census.gov/prod/cen1990/cqc/cqc7.pdf U.S. Census Bureau (1992a). We asked... you told us: Race Washington, DC: Author. Retrieved from: http://www.census.gov/prod/cen1990/cqc/cqc4.pdf U.S. Census Bureau (2011). Overview of race and Hispanic origin: 2010. Washington, DC: Author. Retrieved from: http://www.census.gov/prod/cen2010/briefs/c2010br 02.pdf U.S. Census Bureau (2012). Current Population Survey, 2012 Annual Social and Economic Supplement; Table 1. Educational Attainment of the Population 18 Years and Over, by Age, Sex, Race, and Hispanic Origin: 2012. Washington, DC: Author. Ullman, S. E. (1996a). Do social reactions to sexual assault victims vary by support provider? Violence & Victims, 11 143156. Ullman, S. E. (1996b). Correlates and consequences of adult sexual assault disclosure. Journal of Interpersonal Violence, 11, 554571. Ullman, S. E. (1996c). Social reactions, coping strategies, and self -blame attributions in adj ustment to sexual assault. Psychology of Women Quarterly, 20, 505526. Ullman, S. E. (1996d). Attributions, world assumptions, and recovery from sexual assault. Journal of Child Sexual Abuse, 6, 1 19. Ullman, S. E. (2000). Psychometric characteristics of the Social Reactions Questionnaire. Psychology of Women Quarterly 24, 257271. Ullman, S. E. (2007). Mental health services seeking in sexual assault victims. Women & Therapy, 30, 61 84. Ullman, S. E. (2010). Talking about sexual assault: Society's res ponse to survivors Washington, DC: American Psychological Association. Ullman, S. E., & Brecklin, L. R. (2002a). Sexual assault history, PTSD, and mental health service seeking in a national sample of women. Journal of Community Psychology, 30, 261279. Ullman, S. E., & Brecklin, L. R. (2002b). Sexual assault history and suicidal behavior in a national sample of women. Suicide and Life -Threatening Behavior, 32, 117 130. Ullman, S. E., & Filipas, H. H. (2001a). Correlates of formal and informal support s eeking in sexual assault victims. Journal of Interpersonal Violence, 16(10) 10281047.
147 Ullman, S. E., & Filipas, H. H. (2001b). Predictors of PTSD symptom severity and social reactions in sexual assault victims. Journal of Traumatic Stress, 14, 369389. Ullman, S. E., & Siegal J. M. (1993). Victim -offender relationship and sexual assault. Violence and Victims, 8 121134. Ullman, S. E., & Siegel, J. M. (1995). Sexual assault, social reactions, and physical health. Women's health (Hillsdale, NJ) 1 (4), 2 89. Ullman, S. E., Filipas, H. H., Townsend, S. M., & Starzynski, L. L. (2006). The role of victim offender relationships in womens sexual assault experiences. Journal of Interpersonal Violence, 21 798819. Ullman, S. E., Najdowski, C. J., & Filipas, H H. (2009). Child sexual abuse, post -traumatic stress disorder, : Predictors of revictimization in adult sexual assault survivors. Journal of Child Sexual Abuse 18(4), 367385. Ullman, S. E., Relyea, M., Peter -Hagene, L., & Vasquez, A. L. (2013). Trauma histories, substance use coping, PTSD, and problem substance use among sexual assault victims. Addictive Behaviors 38(6), 2219 2223. Ullman, S. E., Starzynski, L. L., Long, S. M., Mason, G. E., & Long, L. M. (2008). Exploring the relationships of women' s sexual assault disclosure, social reactions, and problem drinking. Journal of Interpersonal Violence 23(9), 12351257. Ullman, S. E., Townsend, S. M., Filipas, H. H., & Starzynski, L. L. (2007). Structural models of the relations of assault severity, s ocial support, avoidance coping, self -blame, and PTSD among sexual assault survivors. Psychology of Women Quarterly, 31, 23 37. United States Department of Justice (2011). Crime in the United States, 2010. Washington, DC: Federal Bureau of Investigation. Urahn, S. K., Currier, E., Elliott, D., Wechsler, L., Wilson, D., & Colbert, D. (2012). Pursuing the American Dream: Economic mobility across generations Washington, DC: Pew Charitable Trust. Urquiza, A. J., & Goodlin Jones, B. L. (1994). Child sexual a buse and adult revictimization with women of color. Violence and Victims, 9 223 232. Van Bruggen, L. K., Runtz, M. G., & Kadlec, H. (2006). Sexual revictimization: The role of sexual self -esteem and dysfunctional sexual behaviors. Child Maltreatment, 11(2), 131145. Van Dijk, J. M. (2001). Attitudes of victims and repeat victims toward the police: Results of the International Crime Victims Survey. In G. Farrell & K. Pease (Eds.), Repeat victimization. Crime prevention studies (Vol. 12, pp. 27 52). Monsey NY: Criminal Justice Press.
148 von Hentig, H. (1940). Remarks on the interaction of perpetrator and victim. Journal of Criminal Law and Criminology, 31, 303309. von Hentig, H. (1948). The criminal and his victim: Studies in the socio-biology of crime. pp. 3 18. Yale University Press; New Haven, USA. Walsh, K., DiLillo, D., Klanecky, A., & McChargue, D. (in press). Posttraumatic Stress Disorder Symptoms: A mechanism in the relationship between early sexual victimization and incapacitated/drug -or alcohol -facilitated and forcible rape. Journal of Interpersonal Violence Walsh, R. M., & Bruce, S. E. (2011). The relationships between perceived levels of control, psychological distress, and legal system variables in a sample of sexual assault survivors. Viol ence Against Women, 17(5), 603618. Weber, M. (1998). Class, status, and party. In R. Levine (Ed.), Social class and stratification: Classic statements and theoretical debates (Second ed., pp. 43 56). Lanham, MD: Rowman and Littlefield. Wenzel, S. L., Ko egel, P., Gelberg, L. (2000). Antecedents of physical and sexual victimization among homeless women: A comparison to homeless men. American Journal of Community Psychology, 28(3), 367390. Wenzel, S. L., Leake, B. D., Gelberg, L. (2000). Health of homeles s women with recent experience of rape. Journal of General Internal Medicine, 15, 265268. Werthem, F. (1949). The show of violence. New York: Doubleday. West, C. M., Williams, L. M., & Siegel, J. A. (2000). Adult sexual revictimization among Black wome n sexually abused in childhood: A prospective examination of serious consequences of abuse. Child Maltreatment, 5 49 57. White, H. R., & Widom, C. S. (2003). Intimate partner violence among abused and neglected children in young adulthood: The mediating effects of early aggression, antisocial personality, hostility and alcohol problems. Aggressive Behavior 29, 332 345. Wilson, A. E., Calhoun, K. S., & Bernat, J. A. (1999). Risk recognition and trauma related symptoms among sexually revictimized women. J ournal of Consulting and Clinical Psychology, 67, 705 710. Wolfgang, M. E. (1958). Patterns in criminal homicide Philadelphia: University of Pennsylvania Press. Wolfgang, M.E. (1957). Victim precipitated criminal homicide. Journal of Criminal Law and Criminology and Political Science, 48, 1 11.
149 Wyatt, G. W., Guthrie, D., & Notgrass, C. M. (1992). Differential effects of womens child sexual abuse and subsequent repeat victimization. Journal of Consulting and Clinical Psychology, 60, 167 173. Zinzow, H. M., Resnick, H. S., Amstadter, A. B., McCauley, J. L., Ruggiero, K. J., & Kilpatrick, D. G. (2010). Drug-or alcohol -facilitated, incapacitated, and forcible rape in relationship to men tal health among a national sample of women. Journal of Interpersonal Violence 25(12), 22172236.
150 APPENDIX A. TABLES Table A. 1: Percentage of Missing Data on Incomplete Cases Variable Names Percent Missing Race/Ethnicity 1.4% Income 12.2% How often in the past week did you feel full of pep? 1.0% How often the past week have you been very nervous? 0.6% How often in the past week have you felt so down in the dumps at nothing to cheer you up? 0.4% How often in the past we could you have a lot of energy? 0.6% How often in the past week did you feel downhearted and blue? 0.6% How often the past week did you feel worn out? 0.4% How often the past week have you been a happy person? 0.4% How often the past week did you feel tired? 0.6% During the past 12 months, how often did you drink any alcoholic beverages, backspace? 0.3% When you were a child did any parent, stepparent, or guardian ever throw something at you that could hurt you? 0.4% When you were a child did any parent, stepparent, or guardian ever push, grab, or shove you? 0.4% When you were a child did any parent, stepparent, or guardian ever pull your hair? 0.8% When you were a child did any parent, stepparent, or guardian ever slapped or hit you? 0.4% When you were a child did any parent, stepparent, or guardian ever kick or bite you? 0.1% When you were a child did any parent, stepparent, or guardian ever choke or attempt to drown you? 0.1% When you were a child did any parent, stepparent, or guardian ever hit you with some object? 0.4% When you were a child did any parent, stepparent, or guardian ever beat you up? 0.1% When you were a child did any parent, stepparent, or guardian ever threaten you with a gun? 0.1% When you were a child did any parent, stepparen t, or guardian ever threaten you with a knife or other weapon besides a gun? 0.1% When you were a child did any parent, stepparent, or guardian ever use a gun on you? 0.1% When you were a child did any parent, stepparent, or guardian ever used a knife or other weapon on you besides a gun? 0.1% Age 1.1%
151 Variable Names Percent Missing Education 0.1% Chronic Injury or Health Problem 0.6% First Child Age 2.6% Domestic Violence Experience 0.4% Victim offender relationship 0.07% Recency of First Rape Victimization 7.9% Pre Victimization Context 33% Was he/she using drugs or alcohol the time of this incident? 20.4% Were you using drugs or alcohol at the time of this incident? 2% Offender Coercion Tactic 0.4% Disclosure to a Therapist 0.6% Informal Support Disclosure 2.1% Time off from School or Work 1.6% Police Engagement 0.9%
152 Table A. 2 : Logistic Regression for Likelihood of Repeat Victimization Model Variable Type Variable B ( SE ) AOR Control Variables Years since rape victimization by first or only perpetrator 0.01(0.01) 1.01 Age at rape victimization by first perpetrator 0.05(0.02)** 0.96 Childhood development years of age) -Parenthood at < 18 years of age 0.11(0.26) 1.12 No children 0.42(0.34) 1.52 Childhood maltreatment 0.49(0.47) 1.64 First perpetrator victimization context Victim offender relationship (reference: Family) --Stranger 1.32(0.43)** 0.27 Acquaintance 0.99(0.29)*** 0.37 Current or Ex Partners 0.22(0.35) 0.80 Hanging out prior to victimization 0.94(0.41)* 2.56 Drinking/drugs prior to victimization 0.13(0.57) 0.88 Date prior to victimization 0.07(0.45) 1.07 Fighting prior to victimization 0.21(0.49) 0.81 Offender use of verbal or physical coercion 0.28(0.12)* 1.32 Whether victim got pregnant from the assault 2.33(1.03)* 0.10 Social power Race/ethnicity 0.08(0.24) 1.09 Income 0.09(0.10) 0.92 Education 0.09(0.10) 1.10 Disclosure of victimization by first perpetrator to therapist 0.06(0.22) 0.94 Disclosure of victimization by first perpetrator to informal sources of support 0.13(0.14) 0.88 Degree of police involvement regarding victimization by first perpetrator 0.14(0.18) 0.87 Lambda 0.18(0.04)*** 0.84 Constant 1.31(0.72) + 0.27 Notes: R 2 = 0.17 (Hosmer & Lemeshow), 0.10 (Cox & Snell), 0.22 (Nagelkerke). Model 2 = 150.36***. + ** ***
153 Table A. 3 : Pre -Victimization with Cross-Tabulation Tables Table A. 3 a : Pre -Victimization Context by Victim Offender Relationship Doing Nothing No Yes Total Stranger 172 26 198 % Within Stranger 86.9% 13.1% 100% Standardized residual 0.2 0.4 Acquaintance 528 82 609 % Within Partner/Acquaintance 86.7% 13.3% 100% Standardized residual 0.3 0.8 Partner 255 26 281 % Within Partner/Acquaintance 90.7% 9.3% 100% Standardized residual 0.5 1.5 Family 258 35 293 % Within Family 88.1% 11.9% 100% Standardized residual 0.0 0.1 1212 169 1381 Notes: 2 = 7.72, p = .17 Table A. 3 b : Pre -Victimization Context by Pregnancy as a Result of First Victimization Doing Nothing No Yes Total No pregnancy 1158 157 1315 % Within No pregnancy 88.1% 11.9% 100% Standardized residual 0.1 0.3 Pregnancy 55 11 66 % Within Pregnancy 83.3% 16.7% 100% Standardized residual 0.4 1.2 2 = 2.22, p = .19 1212 169 1381