ADULT MALES WITH CHILDHOOD SEXUAL EXPERIENCES: THE ROLE OF ATTACHMENT AND COPING IN PSYCHOLOGICAL ADJUSTMENT Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. _________________________ Patricia N. Lyle Certificate of Approval: _________________________ _________________________ Frank W. Weathers, III Barry R. Burkhart, Chairman Associate Professor Professor Psychology Psychology _________________________ _________________________ Alejandro A. Lazarte Martha C. Escobar Assistant Professor Assistant Professor Psychology Psychology _________________________ Stephen L. McFarland Acting Dean Graduate School ADULT MALES WITH CHILDHOOD SEXUAL EXPERIENCES: THE ROLE OF ATTACHMENT AND COPING IN PSYCHOLOGICAL ADJUSTMENT Patricia N. Lyle A Dissertation Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Auburn, Alabama December 15, 2006 iii ADULT MALES WITH CHILDHOOD SEXUAL EXPERIENCES: THE ROLE OF ATTACHMENT AND COPING IN PSYCHOLOGICAL ADJUSTMENT Patricia N. Lyle Permission is granted to Auburn University to make copies of this dissertation at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. ________________________ Signature of Author ________________________ Date of Graduation iv VITA Patricia N. Lyle was born April 30, 1959 in Tampa, Florida. She graduated with honors from Hillsborough High School in 1977 and subsequently joined the United States Air Force as a musician. After 20 years of exemplary service, and an Associate of Applied Science degree in Music, she retired as a Senior Master Sergeant in 1997 and began her current academic career in 1998. She graduated summa cum laude from Alabama State University, Montgomery, Alabama in May 2000 with a Bachelor of Science degree in Psychology. She earned her Masters of Science degree in psychology at Auburn University in 2003 as part of a doctoral program in Clinical Psychology. During her matriculation at Auburn University, her therapy practicum assignments have included one year in the Auburn University Psychological Services Center, two years in the Alabama ? Auburn Consortium?s ABSOP II Juvenile Sex Offender Treatment Program, and one year as the Student Services Counselor for the Auburn University College of Veterinary Medicine. Her psychology residency assignment is in Bay Pines Veterans? Hospital, Bay Pines Florida. She lives with her husband Mike of 23 years, and has two children 22-year-old daughter, Samantha, and 20-year-old son, Nate. v DISSERTATION ABSTRACT ADULT MALES WITH CHILDHOOD SEXUAL EXPERIENCES: THE ROLE OF ATTACHMENT AND COPING IN PSYCHOLOGICAL ADJUSTMENT Patricia N. Lyle Doctor of Philosophy, December 15, 2006 (M.S., Auburn University, 2003) (B.S., Alabama State University, 2000) 205 Typed Pages Directed by Barry R. Burkhart, Ph.D. Research into the sexual abuse of children has focused primarily on female victims, although in the past three decades, the existing literature pertaining to the sexual abuse of boys has expanded. Much of the current male sexual abuse research is directed toward understanding how abused boys later become offenders. While this focus is essential in developing models of abuse ? offender outcomes, these models do less to explain the non-offender outcomes that certainly must exist. By examining the literature pertaining to sexual abuse, sexual offending, and coping with trauma, models were developed to represent how abused boys might successfully process trauma. At the core of such models is the hypothesis that it is not the presence of sexualized coping, but the absence of adaptive coping that distinguishes offenders from non-offenders. vi One hundred sixty-one college men completed anonymous surveys detailing their sexual histories; attachment relationships with mother, father, and friends; their use of sexual and non-sexual coping strategies to relieve distress; and their current psychological adjustment. Measures included the Childhood Sexual Experiences Checklist, the Experiences in Close Relationships Inventory, the Inventory of Parent and Peer Relationships, the Coping Using Sex Inventory, the Brief-COPE, and the Personality Assessment Screener. Nearly 20% of the men experienced exploitive or abusive sex acts, although only 4% defined themselves as abused. Sixty percent admitted compulsive masturbation at some earlier time, with a sample average masturbation rate of 12 times in the preceding month. Coping behaviors were modeled as orthogonal constructs, with sexual and avoidant coping diverging from problem-focused and emotion-focused coping. Hypotheses linking child sex abuse (CSA), coping, and attachment were supported. Poor paternal attachment predicted exposure to sexual abuse and the use of sexual force; poor maternal attachment predicted poor adaptive coping, reduced interest in consensual sex, and increased use of masturbation. Coping fully mediated the direct effect of abuse on attachment. Hypotheses linking CSA, coping, and adjustment were also supported. Abuse was not directly related to psychological adjustment, but operated through coping (maladaptive or adaptive) to produce divergent effects on adjustment. Men who had been abused were more likely than non-abused men to use sex as a coping strategy, less likely to produce adaptive coping, and more likely to provide poorer adjustment scores at the total score and subscale levels. Composite models of all latent variables indicated attachment quality regulates the available coping responses and affects adjustment. vii ACKNOWLEDGEMENTS The author would like to thank Dr. Barry Burkhart for his patient support and guidance in the development of this dissertation and Dr. Alejandro Lazarte for his expertise in statistical analysis. Dr. Frank Weathers? persistent admonishment to ?keep it simple? may seem to have gone unheeded, but in fact served to trim a monstrously ambitious project into a somewhat manageable one. Dr. Martha Escobar graciously agreed to take on this project upon short notice; her support in accessing research participants is dearly appreciated. Thanks must be extended to Sam Singer, Information Technology Specialist, and his assistant Srinivasa Inaganti for their immeasurable assistance in developing the web design and database management. Thanks also to Pamela Dolbear, research assistant, for her assistance in web survey design and data collection. Although they are unaware of their assistance to this work, the many adolescent sex-offenders in treatment at ABSOP II provided great inspiration in molding this research conceptualization. Finally, thanks to the members of Eclectic United Methodist Church for their unending prayers and support. For every individual mentioned here, none deserves greater thanks than husband, Mike Lyle, who sacrificed ?quantity? for ?quality? time, and often was still short- changed. His unyielding commitment to this academic pursuit cannot be measured or fully appreciated, but there will certainly be opportunities to try. viii Style manual or journal used APA Publication Manual (5 th edition) Computer software used Microsoft Word 2003 ix TABLE OF CONTENTS LIST OF TABLES AND FIGURES ...............................................................................xi I. INTRODUCTION ...............................................................................................1 Background..........................................................................................................4 Scope of Research ....................................................................................4 Epidemiology ...........................................................................................4 Definitional Standards ..............................................................................5 Sexual Offender-specific Research ...........................................................8 The Family Environment ..........................................................................9 Child Sexual Abuse and Attachment .................................................................10 Attachment and Internal Working Models ..............................................12 Coping ...............................................................................................................15 General Coping Research .......................................................................15 Attachment and Coping ..........................................................................22 Coping with Childhood Sexual Abuse ....................................................25 Sexual forms of coping with abuse..............................................27 Child Sexual Abuse, Attachment, and Coping ....................................................30 Hypotheses.............................................................................................33 Preliminary Models ................................................................................36 Mediational Model: CSA ? Attachment ? Coping ? Psychological Adjustment ..37 II. METHOD..........................................................................................................38 Participants.............................................................................................38 Instruments.............................................................................................38 Procedure ...............................................................................................40 Electronic Internet-based research...............................................41 Analyses.................................................................................................43 Hypothesis 1 ..............................................................................45 Hypothesis 2 ..............................................................................47 Hypothesis 3 ..............................................................................47 Hypothesis 4 ...............................................................................48 III. RESULTS..........................................................................................................50 Overview................................................................................................50 Survey Format........................................................................................50 Demographic Characteristics ..................................................................54 Family environment ....................................................................54 Mental health ..............................................................................55 x Sexual History........................................................................................55 Childhood sexual experiences .....................................................58 Childhood Sexual Experiences and Related Variables ............................60 Sexualized coping measures........................................................64 Attachment to Parents and Friends..........................................................65 The Experiences in Close Relationships Inventory......................65 The Inventory of Parent and Peer Attachment .............................67 Coping....................................................................................................72 Exploring Hypothesis 1 ..........................................................................76 Models of attachment and coping................................................76 Hypothesis 1...........................................................................................77 Model of attachment and sexualized coping ................................78 Modeling attachment and coping.................................................81 Attachment qualities predict specific coping behaviors ...............86 Hypothesis 2...........................................................................................91 Hypothesis 3...........................................................................................94 Exploring abuse indicators of attachment and coping ..................98 Mediation of the abuse ? attachment relationship ........................99 Exploring specific predictors of sexualized coping and abuse....102 Summarizing predictors of coping, abuse, and adjustment.........108 Hypothesis 4.........................................................................................110 Consolidation of the Models.................................................................115 IV. DISCUSSION..................................................................................................119 Overview..............................................................................................119 General Findings ..................................................................................122 Sexualized coping .....................................................................124 Attachment ...............................................................................124 Coping ......................................................................................125 Hypothesis one .....................................................................................126 Hypothesis two.....................................................................................132 Hypothesis three...................................................................................134 Hypothesis four ....................................................................................139 Coping ......................................................................................140 Attachment ...............................................................................140 Limitations ...........................................................................................141 Conclusions..........................................................................................144 Summary..............................................................................................148 V. REFERENCES ................................................................................................150 VI. APPENDICES .................................................................................................160 A. Instruments ...............................................................................161 B. Information Letter and Resources..............................................188 xi LIST OF TABLES AND FIGURES TABLES 1. Survey Format ...................................................................................................51 2. Differences by Survey Format............................................................................52 3. Sexual History ...................................................................................................57 4. Correlated Demographic and Sexual Variables within the CSEC........................61 5. Sexualized Coping, Sexual Experiences, and Correlates within the CSEC..........62 6. Correlated Sexual Abuse Variables ....................................................................63 7. Summary Information for Coping Using Sex Inventory (CUSI) .........................64 8. Correlations between Sexual Coping Measures ..................................................65 9. Descriptive Statistics for Experiences in Close Relationships Inventory (ECL) ..66 10. Descriptive Statistics for the Inventory of Parent and Peer Attachment (IPPA)...68 11. Rotated Component Matrix for Situational Coping.............................................73 12. Rotated Component Matrix for Dispositional Coping.........................................74 13. Sexualized Coping and the Situational 5-Factor Coping Model ..........................75 14. Sexualized Coping and the Dispositional 5-Factor Coping Model ......................76 15. Attachment and the Dispositional 5-Factor Coping Model..................................77 16. Model 1 Regression Weights..............................................................................81 17. Model 3 Regression Weights..............................................................................85 18. Sexual and Non-sexual Predictors of Attachment (n=151)..................................88 xii 19. Sexual and Non-sexual Predictors of Attachment within ECL Groups (n=151) ..89 20. Model 5 Regression Estimates ...........................................................................94 21. Abuse Indicators Predicting Attachment Problems, Sex Coping, and Non-sex Coping ..........................................................................................................98 22. Predictors of Sexual Coping, Childhood Abuse, and Adjustment......................103 23. Predictors of Sex Coping and CSA Groups (n=151) .........................................105 24. Predictors of Abuse and Adjustment among the Abused (n=27) .......................106 25. Predictors of Sex Coping and CSA Groups among the Abused (n=27) .............106 26. Predictors of Sex Coping and Adjustment in the Non-abused Group (n=122)...107 27. Predictors of Sex Coping in the Non-abused Group (n=122) ............................108 xiii FIGURES 1. The Impact of Attachment Style.........................................................................37 2. The Impact of Attachment Figure.......................................................................37 3. Survey Format ...................................................................................................51 4. ECL and Compulsive Masturbation....................................................................66 5. 30-day Masturbation and Mother Communication..............................................70 6. 7-day Masturbation and Parent Security.............................................................70 7. Compulsive versus 30-day Masturbation and Mother Communication ...............71 8. Compulsive versus 7-day Masturbation and Parent Attachment..........................71 9. Model 1 ? Attachment and Sexualized Coping, Age-restricted (n=151)..............80 10. Model 2 ? Exploring Attachment and Coping.....................................................82 11. Model 3 ? CSA, Attachment, and Coping...........................................................84 12. Model 4 ? CSA, Attachment, and Coping (Revised Model)................................86 13. Model 5 ? Self-identified Abuse, Attachment, and Coping .................................93 14. Model 6 ? Relationship of CSA to Attachment and Coping................................95 15. Model 7 ? Direct Effects of CSA on Coping and Attachment .............................97 16. Model 8 ? Mediation: CSA, Attachment, and Coping.........................................99 17. Model 9 ? Mediation of the Direct Effect of CSA on Attachment.....................101 18. Model 10 ? CSA, Coping, and Adjustment.......................................................111 19. Model 11 ? Mediation: CSA, Coping, and Adjustment.....................................112 20. Model 12 ? CSA, Coping, and Adjustment.......................................................114 xiv 21. Model 13 ? Mediational Influence of Attachment and Adjustment ...................116 22. Model 14 ? Direct Effects of Coping on Current Attachment and Adjustment..117 1 INTRODUCTION Childhood sexual abuse has been examined from many different perspectives. Early descriptive work attempted to identify common characteristics of abuse, prevalence rates, demographic profiles of victims and perpetrators, and sequelae particular to childhood sexual abuse (Finkelhor, 1979, 1981). Later work focused on understanding the emotional, psychological, and behavioral effects of abuse by producing basic models of abuse ? effect relationships (Finkelhor & Browne, 1985; Fromuth & Burkhart, 1989; Roche, Runtz, & Hunter, 1999). Other work examined the existing research and produced conflicting information about even the most basic assumptions, highlighting the lack of agreement among researchers in defining abuse, measuring important variables of interest, and identifying important abuse ? effect relationships (Madu & Peltzer, 2001; Rind, Tromovitch, & Bauserman, 1998; Stander, Olson, & Merrill, 2002). The current status of the field of child sexual abuse is such that it is possible for researchers to focus upon one population, one syndrome, or one treatment to the exclusion of all others, and in doing so, produce a wealth and variety of information that can at times be difficult to integrate. Since the early 1980s, research has increasingly included the investigation of the sexual abuse of boys. Finkelhor?s The Sexual Abuse of Boys (1981) is often cited as the seminal work in precipitating an understanding of the victimization of this population. 2 Since that time, much of the research into sexually abused boys has focused on those who become sexual offenders in adolescence and adulthood. The emergence of a connection between sexual offending and being sexually victimized was a logical direction for researchers to take in that both areas are important problems in our society, and have been associated through retrospective reports by incarcerated sexual offenders. Unfortunately, in the eagerness to understand this important relationship, researchers have neglected to recognize or examine those boys who have been sexually abused and not become offenders. The current project examines relationships between childhood sexual experiences and later sexual behavior in a sample of non-offending men, attempting to extend and incorporate important concepts from contemporary child abuse and sex-offender research. Of particular interest is the relationship between sexual abuse, attachment, coping styles, and psychological adjustment. A pilot study (Lyle, 2003) found that in a college sample of 107 men, childhood sexual abuse predicted sexualized coping behavior, and that attachment mediated the effect of this relationship. Men in this study who recognized they were victims of childhood sexual abuse reported a number of distressing experiences that were significantly related to abuse. Being forced into sex before age 13 was found to be related to childhood exposure to domestic violence committed by the father, younger age at first interpersonal sexual experience, unwanted sexual arousal, use of counseling and mental health services, the use of frequent masturbation under stress, and pressuring someone for sex. Those men who admitted being forced into sex before age 13 also were more likely to have experienced childhood intercourse with either family members or non-family members. 3 In this same study, attachment anxiety was positively related to sexualized coping (including both masturbatory and interpersonal behavior) while attachment avoidance was not significantly related to these behaviors. Difficulties with parent trust and communication provided the strongest correlations with increased sexualized coping. However, when examining mothers? and fathers? care and overprotection separately, attachment to mothers was found to be a better predictor of sexualized coping in this sample. This study demonstrated that, in part, the relationship between childhood sexual abuse (CSA) and sexualized coping operated through parental attachment. In broad terms, those men with the strongest bonds with their parents were the least likely to use sexual means to deal with stressful experiences. The current project extends these findings to other forms of coping, incorporating knowledge about attachment and gender differences in coping with trauma. Specifically, those men with childhood sexual experiences are expected to use coping strategies whose functionality and adaptability is mediated by the quality of attachment to their parents and friends. Also, men who exhibit sexualized forms of coping are predicted to have fewer coping strategies overall than those who were not abused. Finally, those men who rely primarily on sexualized coping strategies will be more likely to engage in coercive sexual behavior, will have the poorest attachment bonds, and may have the most severe childhood sexual abuse when compared to other groups in the sample. 4 Background Scope of Research The study of childhood sexual abuse has received wide attention over the preceding three decades as a result of public awareness of the detrimental effects of sexual contact between adults and children. To demonstrate the extent of such interest, a simple search was conducted in several scientific databases for the terms child sexual abuse. Nearly two thousand articles appeared in Infotrac dating back to 1982, while only 63 of those pertained to boys. In ERIC, the same search produced over eighteen thousand articles with the earliest dated 1964, while only 10 pertained to boys. A search of OVID produced 978 articles dating back to 1978 with only 41 pertaining to boys. Searching PsycINFO produced 2879 articles related to child sexual abuse with 77 pertaining to boys. The earliest article found was from the year 1913 and referred to the psychology of victims? testimony and support in alleged cases of child sexual abuse (Whipple, 1913). This series of searches outlines the intense interest placed upon child sexual abuse, the recent timing of most investigations, and the relatively limited proportion of interest in the sexual abuse of boys. Epidemiology Research into child sexual abuse has produced widely differing prevalence rates. Representative of such work, Madu and Peltzer (2001) examined a broad range of studies yielding prevalence rates for women from 7% (Siegel, Sorenson, Golding, Burman, & Stein; 1987) to 62% (Wyatt, 1985), and for men from 4% (Siegel et al., 1987) to 30% (Landis, 1956). Some of the disagreement in these figures is a result of differing 5 definitions of what types of acts constitute abuse (contact vs. non-contact), what age of contact (and difference in age of perpetrator and victim) signifies childhood abuse, and what type of outcome criteria constitute trauma or maladjustment related to the abuse. In 1982, the Federal Bureau of Investigations? Uniform Crime Report stated that one in three women and one in ten men had been sexually assaulted by someone they knew by age 13. A recent study of over 11,000 male and female Navy recruits (Stander, Olson, & Merrill, 2002) found that 47% of women and 23% of men reported childhood sexual experiences (CSEs) prior to age 18 with someone at least 5 years older. Furthermore, 18% of women and 3% of men in this study reported CSEs with an immediate or extended family member at least 5 years older. The mean age of contact for those in the non-familial CSE group was just under 12 years old. For the familial CSE group, the mean age of first sexual experience was approximately 10 years old. Stander et al. discovered that in the non-familial abuse group, women (49%) identified themselves as sexual abuse victims more often than did men (15%), even though further questioning using commonly used operational definitions of abuse revealed higher numbers for men. Definitional Standards While there has been much interest in the consequences of child sexual abuse, there has been disagreement on what constitutes abuse for boys. For example, in a study of 582 college men, Fromuth and Burkhart (1987) reported that, depending upon the criteria of the definition of childhood sexual abuse, 4% to 24% could be classified as abused. When the definition was restricted to experiences in which (a) the child was less 6 than 13 years old, (b) the perpetrator was 5 years older and at least 16, and (c) the subject viewed the experience negatively, then only 4% of men this classification. Stander et al. (2002) found that men were less likely to acknowledge abuse and to report CSE characteristics indicative of abuse, except for CSEs in which the perpetrators were also male. In 1981, Finkelhor reported that although it was expected in early research to find that boys were abused mostly by women, it appeared instead that men were abusing the majority of boys as well as girls. While sexual abuse of boys by older males seems to be more easily recognized by the victim and society as sexual abuse, it is apparently nonetheless underreported. Finkelhor (1981) suggested that the underreporting of child sexual abuse of boys was related to aspects of shame, denial, and fear on the part of male victims much the same as with abused girls. However, in the case of boys, Finkelhor also suggested that abuse at the hands of a male places the boy in the uncomfortable and unacceptable position of admitting he does not meet the stereotypical male role model of being confident, strong, and independent. Needing help because one has been coerced or forced into a taboo relationship produces considerable strain and reduces the likelihood a boy will feel justified asking for help. Additionally, the assumptions about his sexual orientation may prevent either he or his family from making such contact known to the authorities (Finkelhor, 1981). In the case of boys victimized by women, the issues are, likewise, more complicated, but in different ways. For example, Stander et al. (2002) found that men who had childhood sexual experiences with females that could be defined as coercive or exploitive often did not define such contact as abusive. They more readily defined acts perpetrated by males as 7 child sexual abuse. Additionally, Fromuth and Burkhart (1989) examined childhood sexual abuse in two samples of college-age men, inquiring about several variables including gender of the perpetrator and the victims? perception of the severity of the incident(s). What they found was surprising in light of previous research that indicated abusers were typically males. In Fromuth and Burkhart?s midwestern sample, 78% of perpetrators were female, while in the southeastern sample, 72% were female. In both samples, the abuse experiences with women were usually not viewed negatively. Collapsing across both samples, 60% of the men reported experiencing interest or pleasure, 28% surprise, and 12% fear or shock. Fromuth and Burkhart suggested that experiences that are consistent with the ?male image? allow men to deny the abusive aspects of those experiences. Duncan and Williams (1998) examined gender-role socialization of males and the impact of male-on-male versus female-on-male sexual abuse. They reported that among those males who were abused, 57% of the sample reported abuse at the hands of a male, while 51% reported abuse at the hands of a female; 73% of the perpetrators were friends or acquaintances. Of the boys whose abuser was female, 81% were acquaintances or friends of the boy, often older female friends or babysitters. Only 9% of their female abusers were family members. In many of the cases of coercive abuse by older females (babysitters and friends), the boy claimed he was the one to initiate sexual contact as a result of the girl showing sexual interest in him. These cases almost always involved sexual intercourse with a female more than 5 years older; however, as with the Fromuth and Burkhart (1989) study, the boy usually did not consider the contact abusive. 8 In an epidemiological study of male childhood sexual abuse, Burkhart, Fromuth, Gold, & Torquato (1993) found that while 27% of respondents reported at least one experience of childhood sexual contact, 43% of the experiences were rated positively, 38% neutrally, and 19% negatively. A generally consistent finding among men who retrospectively report childhood sexual experiences is that more of them, compared to women, do not find these incidents distressing. The tendency for men to view early sexual experiences as non-abusive or neutral means that researchers must re-examine the parameters of abuse and its outcomes for men. The lack of subjective negative experience in boys? understanding of some forms of childhood sexual contact has implications for further research. If boys do not recognize certain forms of sexual contact as abusive, are they nonetheless likely to manifest psychological or social outcomes that are pertinent to their adult functioning? Sexual Offender-specific Research At first glance, the predominant focus of research into the sexual abuse of boys is whether an abused boy is likely to become a sexual offender (Marshall & Marshall, 2000; Merrill, Thomsen, Gold, and Milner, 2001; Murray, 2000; Smallbone &McCabe, 2003; Smallbone & Wortley, 2000). Smallbone and Wortley studied a large population of incarcerated offenders and found that a large proportion of men (from 55% to 73%, depending on offense type) reported having experienced childhood sexual abuse before becoming offenders. Murray (2000) reported that in a study of acknowledged pedophiles and hebephiles (those who prefer adolescents between ages 13 and 16), 42% and 44%, respectively, claimed to have been sexually abused during childhood. Merrill et al. (2001) 9 found that men in their study who had experienced both physical and sexual abuse during childhood were 400-600% more likely to have committed rape as their non-abused counterparts. Marshall and Marshall (2000) proposed a model for the development of sexual offending behavior in which insecure attachment bonds during childhood place the child at greater risk for being sexually abused. The quality of the parental attachment then influences the processing of that abuse and later development of sexual offending behavior. In a sense, the family environment has a key role in creating both physical and psychological vulnerability for abuse and in mediating the effects of that abuse in later life. The Family Environment Some researchers suggest that family environment, not sexual abuse, per se, more directly influences maladaptive functioning in later life (see Merrill, Thomsen, Sinclair, Gold, & Milner, 2001). In their 1998 meta-analysis of CSA research using college samples, Rind, Tromovitch, and Bauserman found that college students with a history of CSA came from more problematic families compared to controls. When they examined the relationship between family environment and symptoms, Rind et al. found that effect sizes indicated that family environment was a stronger predictor of most symptomatology than was childhood sexual abuse. There is general agreement that severity of abuse (i.e. sexual contact, penetration, relation to the abuser, and number of occurrences) influences the severity of later maladaptive functioning, but in many cases, family dysfunction accounts for more of the variance in adjustment and behavior than the abuse-specific variables (Alexander, 1992). 10 Child Sexual Abuse and Attachment The literature linking the short- and long-term effects of sexual trauma to attachment is steadily increasing. For example, Roche, Runtz, and Hunter (1999) found in their female sample that a history of child sexual abuse predicted both psychological adjustment and adult attachment style, and that adult attachment style mediated the abuse ? psychological adjustment relationship. Roche et al. estimated that between 70% and 100% of maltreated children exhibited insecure attachment patterns versus a rate of about 30% in the general population. Since attachment is seen as fundamental to the development of the self, insecure attachment leaves a child unable to protect the self from his outward experience of others. In cases of sexual abuse, this inability to buffer against assault to the self (and the person?s sexual being) leads to psychological maladjustment. In Roche et al., abuse occurring within the family also was found to be more detrimental than extrafamilial abuse. The most important impact of CSA found in this study was the view of oneself as undeserving of love and support from others. Alexander (1992) suggested that long-term effects of childhood sexual abuse could be understood in terms of the attachment relationships available at the time of the abuse. Moreover, the survivor?s adult attachment style will influence the expression of long-term effects. For example, the preoccupied individual may be more likely to manifest a desperate or manic love style that often results in disappointment and further revictimization. The avoidant individual would likely experience social isolation, dependency, and lack of trust that ultimately manifests as compulsive sexuality without emotional intimacy (Alexander, 1992). 11 There is increasing interest in examining the impact of attachment in the resolution of boys? sexual experiences. Several researchers have proposed that the quality of attachment in a young boy?s life impacts his exposure to early sexual experiences (Marshall & Marshall, 2000), influences his ability to cope with the aftermath of such experiences (Smallbone &McCabe, 2003), and exerts some control in the development of sexually coercive or offending behavior (Marshall & Marshall, 2000; Smallbone & Dadds, 1998, 2000; Smallbone &McCabe, 2003). Furthermore, some researchers (Marshall & Marshall, 2000; Smallbone & Dadds, 1998; 2000) have suggested that it is the quality of attachment to the father that functions most directly in the development of sexual offending against children. This basic assumption relating CSA, attachment, and sexualized coping behavior was examined in a study of 107 college men (Lyle, 2003). Approximately 5% of the men identified themselves as having been sexually abused as children. When abuse was defined as sexual experiences with a family member, those that included force, or those with a person 5 years older, the rate of CSA increased to over 20%. In the study, attachment to parents was significantly related to sexualized coping behavior including frequent masturbation and using interpersonal sexual behavior to relieve stress. Furthermore, there was a significant relationship between CSA and sexualized coping that was reduced to nonsignificance when attachment was introduced into the model. When attachment to parents was examined separately, attachment to mother was more useful in understanding the non-deviant forms of sexualized behavior while attachment to father was more important in understanding more deviant behavior, including 12 masturbating to fantasies of forced sex with children. Finally, poor attachment with peers was significantly related to coercive sexual behavior with adults when under stress. Attachment and Internal Working Models Attachment is the process of achieving bonds with important caregivers for the purpose of safety and security. In the ethological view of attachment, efforts to maintain caregiver proximity have evolved in our species because they promote survival (Berk, 1998). Bowlby (1988) proposed that proximity-seeking behavior was not designed solely for feeding purposes but also for protection and comfort. A distressed child will intensify his attempts to gain comfort and protection from his caregiver, and the type of response he receives during such times provides him with information about caregiver consistency, availability of protection, and his own self-worth. The child develops an internal working model about the availability of attachment figures and their likelihood of providing support and comfort during stressful experiences. The internal working model is derived from past experiences with caregiver consistency and quality of protection and guides his expectations for future support and protection, as well as his interpretation about potential risks in threatening situations (Berk, 1998). Children whose parents respond supportively to distress will develop internal working models that represent the world as a safe place where needs will be met and caregivers can be relied upon for support and protection. These children explore their environment in relative comfort and return to the caregiver as a secure base. A child whose parent responds inconsistently develops a disorganized pattern of attachment in which his internal working model represents the world as an unsafe and unpredictable place where needs may not be met. Children in this situation 13 often intensify attachment behaviors becoming clingy, yet inconsolable, in their distress. When a parent consistently responds to a child?s distress with anger, violence, and neglect, the child develops an internal working model of the world as a place where needs often will not be met, the child is unworthy of protection, and others are dangerous. Thus, internal working models can be seen to play a key role in how a child or adult approaches his environment (Thomas, 2000) and resolves distress. According to Berk (1998) and others, childhood attachment patterns are good indicators of adult internal working models and relationship experiences. Berk summarizes the work of a number of attachment researchers in studying adult romantic relationships. Securely attached adults see themselves as likeable and easy to get along with, are comfortable with intimacy, and are rarely worried about abandonment or getting too close to someone. Their romantic relationships are described as involving trust, happiness, and friendship. Adults with avoidant attachment patterns have learned from demanding, critical, and disrespectful parents that the world is predictably unsafe. They have internal working models that stress independence, mistrust of their partners, and fear about getting too close. They are convinced that others dislike them and romantic love is hard to find and rarely lasts. Adults with resistant attachment had parents who responded unpredictably and unfairly and, therefore, have developed internal working models that compel them to merge completely with another person and fall in love quickly. At the same time, they worry that their intense feelings will overwhelm others who would eventually abandon them. These relationships are marked by jealousy, emotional lability, and worries about whether the partner will return their affection (Berk). 14 Thomas (2000) proposed a theory of the long-term effects of child abuse that emphasized the relationship between inadequate caregiver protection and internal representations of an effective protector. He theorized that children who are poorly protected have ongoing difficulty protecting themselves from interpersonal aggression and have trouble with internal self-criticism. They often have confused and negative representations of themselves and their parents and have poorer relationships with their peers throughout life. According to Thomas (2000), maltreated children have developed internal representations of themselves in relation to their caregivers and internalize three roles: victim, victimizer, and protector. Children may identify with their representations of their caregiver and later use these representations as templates for self-concept and behavior. Those with inadequate protectors may develop internal working models of the endangered child, an inadequate inner protector, and an uncontrolled aggressor. Thomas suggests that the role of endangered child correlates to adult symptoms such as anxiety, phobias, hypervigilance, chronic insecurity, mistrust, betrayal, and need for control. The role of inadequate inner protector produces passivity, failure to maintain boundaries, disregard for personal safety, and revictimization. The role of uncontrolled aggressor is related to self-injury, suicide, victimization of others, and an exaggerated sense of personal evil. Thomas believes that depression among abuse survivors is an interaction of all three roles, where an inner dialogue or ?drama? occurs between an abusive critic and a passive child personality. It may be that such negative self-talk limits the coping behaviors a person believes are viable options during times of stress. 15 Coping General Coping Research Coping is a transaction between a threat, an appraisal of that threat, and a response (Tamres, Janicki, and Helgesen, 2002). Much of the research into coping recognizes two broad types of coping ? problem-focused and emotion-focused. Problem- focused coping is aimed at problem solving or doing something to alter the source of stress. Emotion-focused coping is aimed at reducing or managing the emotional distress that is associated with the situation (Lazarus & Folkman, 1984; Carver, Scheier, & Weintraub, 1989). Problem-focused coping implies that the person believes that something constructive can be done, while emotion-focused coping tends to occur when people feel that the stressor is something that must be endured (Carver et al., 1989). Examples of emotion-focused behaviors include venting, ruminating, avoidance, acceptance, interpreting the problem in a positive light, and self-blame (Tamres et al., 2002). Problem-focused strategies include active planning, suppression of competing activities, exercising restraint, and seeking instrumental support. Coping involves learned behaviors that contribute to survival in the face of life- threatening dangers (Folkman & Lazarus, 1988). Coping behaviors are initiated by fear and anger and result in avoidance or escape in the former, and confrontation or attack in the latter. Coping also includes cognitive processes such as denial, repression, suppression, and intellectualization. Problem-solving behaviors are invoked to reduce or manage anxiety and other distressing emotions. Because coping is a dynamic process between behavioral and emotional strategies, the coping strategies that are helpful will be 16 specific to the actual traumatic event and will depend on the victim?s appraisal of the event, time since trauma, personality, and available social support. Rothbaum, Weisz, and Snyder (1982) describe coping in terms of primary and secondary control theory. Primary control refers to efforts to change existing realities and secondary control refers to attempts to accommodate to those realities. In this basic view, people try to first change a stressful situation until their efforts fail, at which time they resign themselves to accept it. However, Bandura (1997) suggests that coping is a higher order process that extends beyond the basic divisions between primary control and secondary control. Primary and secondary control models fail to recognize that a response to a stressor may change that stressor or affect the existing emotional response to it. In turn, the individual?s perception of the changed problem or emotion impacts his continued efforts to change it or accommodate it. According to Bandura, the ongoing reciprocal influence of continuous assessment is an important part of the coping process. Further, he maintains that self-efficacy beliefs exert important control in the type of strategy one uses to adapt to a particular stressor. Self-efficacy beliefs pertain to one?s ability to produce an effective response to a particular stressor. These beliefs are derived from one?s repertoire of available responses, a history of successful use of a particular response, and the general appraisal of how closely the current stressor matches previous situations in which the response was successful. Coping efficacy beliefs are impacted by negative thoughts and emotions, risk discernment, and beliefs about personal vulnerability (Bandura, 1997). The interplay between various contributors to efficacy beliefs makes this process dynamic and fluid. Under stressful situations, cognitive schemas are activated and influence the perception 17 of threat, the belief in one?s ability to respond, and the choice of responses. The resulting reciprocal cycle of success or failure impacts those schemas and influences future responding. People tend to adopt certain coping tactics as relatively stable preferences, which may derive from personality or other stable factors. Situational factors that tend to impact coping preferences revolve around the perceived controllability of the stressor and the relative importance of the outcome (Carver et al., 1989). According to Tamres et al. (2002), dispositional characteristics (psychological, emotional, and biological) impact coping choices at certain times while situational differences are more influential at other times. Additionally, societal roles men and women assume and the subsequent stressors men and women face often produce sex differences in coping (Tamres et al., 2002; Kimerling, Ouimette, & Wolfe, 2002). Optimism also appears to impact situational responses to trauma. Carver et al. (1989) define optimism as the expectation of favorable outcomes that promotes the use of active coping, while pessimism implies unfavorable expectations and implies dealing with emotional distress and disengagement. When situations are controllable, active coping strategies are more likely to be used. When situations seem less controllable, alternative strategies predominate. Coping research has examined responses to varied stressors including cancer treatment (Fillion, Kovacs, Gagnon, & Endler, 2002; Boman & Bodegard, 2000; Hampton & Frombach, 2000), psoriasis (Hill & Kennedy, 2002), depression (Wilhelm, Roy, Mitchell, Brownhill, & Parker, 2002), bereavement (Nolen-Hoeksema, Parker, & Larson, 1994), HIV/AIDS (Simoni & Ng, 2000), and multiple forms of childhood trauma 18 (Lev-Wiezel, 1999; Ballon, Courbasson, & Smith, 2001; Swanston, Nunn, Oates, Tebbutt, & O?Toole, 1999; Skinner, 1999; and Elmone & Lingg, 1996). Additionally, researchers have sought to delineate specific coping styles and characteristic coping responses within those styles. While early theories focused on problem- versus emotion- focused strategies (Lazarus & Folkman, 1984) or primary and secondary control strategies (Rothbaum et al., 1982), as coping measures were developed, researchers often found coping could be described in varied and more complex ways (Carver et al., 1989). A brief review of coping literature revealed many different measures of coping in varied populations. Merrill, Guimond, Thomsen, and Milner (2003) investigated the role of coping style in female Navy recruits who experienced childhood sexual abuse. They used a modified version of the How I Deal with Things Scale by Burt and Katz (1987) in which participants rated 30 statements related to how they responded to their sexual abuse in the weeks following the incident. Subscales derived from these responses included self-destructive coping and avoidant coping. Coffey, Leitenberg, Henning, Turner, and Bennett (1996) used the Coping Strategies Inventory to assess how a community sample of women coped in adulthood with childhood sexual abuse. The measure (CSI; Tobin, Holroyd, Reynolds, & Wigal, 1989) used 76 items to produce subscales related to engagement and disengagement. Simoni and Ng (2000) used a modified version of Folkman and Lazarus?s Ways of Coping Scale (1986) to examine coping in a sample of women with HIV/AIDS. Their version of the scale included a list of 33 coping behaviors specific to dealing with HIV/AIDS and produced subscales related to avoidant coping and adaptive coping. Lev-Wiesel (1999) used the Potency Scale (Ben-Sira, 1985) to examine adults coping with child abuse by their parents. This 19 scale included 19 items related to self-confidence, feelings of control, feelings of social obligation, and society perceived as ordered and meaningful. Ballon et al. (2001) assessed youths using drugs and alcohol to cope with physical and sexual abuse. Their measure of coping involved a clinical interview to determine the frequency and duration of substance use relative to childhood abuse. Of the coping measures reviewed, Carver et al.?s COPE (1989) provides the most comprehensive, theory-based method of assessing and describing coping behaviors, and appears to be useful across widely varied stressors. Sexualized coping strategies have been of recent interest in the sexual trauma literature (Duncan & Williams, 1998; Hall, Mathews, & Pearce, 1998; Gartner, 1999); particularly the use of masturbation and interpersonal sexual contact to relieve stress. In the sex-offender specific literature, several researchers (Marshall & Marshall, 2000; Smallbone &McCabe, 2003; Cortoni & Marshall, 2001) have suggested that sexual behavior may serve as a specific coping strategy for dealing with stress, and precipitates the appearance of sexual offending. Cortoni and Marshall developed a 16-item scale to assess the use of sexual strategies to relieve distress. These items produced three factors related to themes of consensual adult sex, forced adult sex, or sex with children. Their Coping Using Sex Inventory distinguished between sexual offenders and nonsexual offenders in a prison population with 74% accuracy. Furthermore, researchers who propose that sexual behavior serves as a coping strategy have linked this strategy to childhood sexual abuse and adolescent sexual preoccupation (Smallbone &McCabe, 2003; Marshall & Marshall, 2000; Duncan & Williams, 1998). However, sexualized coping may be ?normal? behavior for males and not specific to sex offenders. 20 Scientific examination of coping is extremely difficult in light of the variability in describing, measuring, or comparing coping strategies. Carver et al. (1989) reviewed existing coping measures and identified three basic problems. First, none of the measures tapped into all areas they believed were theoretically relevant. Second, the scales included items that were lacking in clarity or focus. Third, most of the scales were empirically derived from a priori statements of coping behaviors that had little or no relation to theory. In response to this evaluation, Carver et al. produced a theoretically and empirically based coping measure that was derived from the Lazarus model of stress and a model of behavioral self-regulation from their earlier research (Carver & Scheier, 1981, 1983, 1985; Scheier & Carver, 1988). The Coping Orientations to Problems Experienced Inventory (COPE; Carver et al., 1989) was developed through a series of studies to assess the different ways people respond to stress. Carver et al. produced 60 scale items resulting in 11 conceptually distinct scales within the COPE: Active Coping and Planning, Suppression of Competing Activities, Restraint, Seeking Social Support, Focusing on and Venting of Emotions, Behavioral Disengagement, Mental Disengagement, Positive Reappraisal, Denial, Acceptance, and Turning to Religion. Cronbach?s alpha reliability coefficients for these scales were above .6 for all but the Mental Disengagement scale, which appeared to consist of several distinct behaviors. When Carver et al. (1989) assessed coping strategies in terms of personality they found that Active Coping and Planning were positively correlated with optimism, the feeling of being generally able to do something about stressful situations, self-esteem, hardiness, and Type A personality. Active coping was inversely associated with trait 21 anxiety. The Denial and Behavioral Disengagement scales were positively correlated with trait anxiety and negatively correlated with optimism, the feeling of being generally able to do something about a stressful situation, and self-esteem. Focusing On and Venting of Emotions was negatively related to the feeling of being able to do something about stressful situations and with internal locus of control, and was positively related to trait anxiety and monitoring. Active Coping, Planning, Suppression of Competing Activities, and Seeking Out Instrumental Support occurred more in controllable than uncontrollable situations. Also, the more the situation mattered to the subject, the more likely the use of venting emotions, engaging in denial, and seeking out social support. Carver et al. (1989) found several sex differences in examining coping responses men and women use when responding to typical stressors. In the dispositional version of the COPE, women more than men reported that they usually sought support for both emotional and instrumental reasons and that they usually focused on and vented emotions. Men reported usually turning to alcohol more than did women. In the situational version of the COPE, differences arose as well. Again, men reported more alcohol use than did women, and women reported seeking social support for emotional reasons more than did men. Finch, Panter, and Caskie (1999) provided an interbattery factor analysis of the revised Ways of Coping Checklist (Folkman, Lazarus, Dunkel-Schetter, Delongis, & Gruen, 1986) and the COPE Inventory (Carver et al., 1989) and achieved a five factor solution. The interbattery method of factor analysis allowed the separation of method- based variability from trait-based variability. Finch et al. identified five distinct dimensions common to both measures of coping. The dimensions included problem- 22 solving, support-seeking, avoidance, reframing, and distancing. Problem-solving included taking direct action to remove the source of stress or reduce its effects. Support-seeking included turning to others for advice, comfort, or aid, and venting one?s feelings. Avoidance included maladaptive strategies of withdrawal, denial, disengagement, use of substances, and self-blame. Reframing included strategies such as focusing on positive personal growth as an effect of the stressor or turning to religion as the basis for positive reinterpretation of the situation. Distancing included accepting the reality of the situation and attempting to detach from it both physically and emotionally. Attachment and Coping Several researchers have established relationships between attachment styles and coping behavior (Creasey & Hesson-McInnis, 2001; Kemp & Neimeyer, 1999; Lopez & Gormley, 2002; Lopez, Mitchell, & Gormley, 2002; Mikulincer, 1998; Mikulincer, Florian, & Weller, 1993; Pierce & Lydon, 1998; Wei, Heppner, & Mallinckrodt, 2003). Mikulincer et al. studied the association between adult attachment style and the way people reacted to missile attack during the Gulf War and found 2 weeks after the attack that secure people used more support-seeking strategies in coping with the attack, while ambivalent people used more emotion-focused strategies, and avoidant people used more distancing strategies. However, the three attachment styles did not differ in the use of problem-focused strategies, which the authors attribute to mass media information about what to do in the event of an attack. Mikulincer et al.?s mediational model did not support the existing assumption that attachment styles produce different coping styles, but that attachment style has a direct effect on both coping and distress. They argue that 23 differences in self-efficacy, control, optimism, and trust may influence the effects of attachment on coping and distress. In 1998, Mikulincer examined the effect of attachment styles on coping with a violation of trust, suggesting that the basic guidelines of the attachment system ? acknowledgement of distress, engagement in constructive actions, and turning to others for support (Bowlby, 1988; Mikulincer & Florian, 1998) ? exert some control in coping behavior. Anxious-ambivalent persons would be more likely to direct attention to distress in a hypervigilant manner and ruminate on negative thoughts. Avoidant persons would likely tend to detach from distressing situations and attempt to achieve cognitive and behavioral distancing from distress cues. Mikulincer found when subjects? trust had been violated by a partner, secure people were the most likely to deal with the event by talking with the partner, anxious-ambivalent people tended to react with ruminative worry, and avoidant people tended to distance themselves from their partner. When Pierce and Lydon (1998) tested the effects of interpersonal expectations on responses to a hypothetical stressful event (unplanned pregnancy), they found that activation of positive interpersonal expectations increased reports of support-seeking behavior and decreased the use of self-denigrating coping. Activation of negative interpersonal expectations decreased growth-oriented coping. Avoidance in relationships was significantly related to self-denigration but anxiety in relationships was not. Neither anxious nor avoidant attachment were associated with growth-oriented coping. Lopez and Gormley (2002) examined college students? stability or change in attachment style and the effect on self-confidence, problem coping styles, and distress. Their research indicated a moderate level of attachment stability during students? first 24 year of college (October and April test dates during one academic year). Participants who changed from secure to insecure reported increased reactive coping over time while stable secure peers showed decreased reactive coping over time. Only the stable insecure group showed increased levels of suppressive coping over time. Mediational models have produced mixed findings in linking the effects of attachment and coping on distress. For example, Creasey & Hesson-McInnis (2001) developed a path analytic model examining the associations between attachment orientations and coping with conflict in the romantic relationships of 357 college students. They proposed that, in distressing romantic situations, anxious participants would be more likely to have difficulty regulating emotions while avoidant participants would have more difficulty displaying adaptive coping strategies. Their model provided limited support for this assumption, in that anxious participants engaged in more conflict negativity and avoidant participants engaged in more withdrawal, but not to a significant degree. Similarly, Lopez, Mitchell, & Gormley (2002) found in a smaller college sample that attachment anxiety was significantly related to reactive and suppressive coping, but avoidance was not related with these coping styles to a significant degree. However, a mediational model predicting student distress from measures of negative life events, adult attachment orientations, and measures of self-organization and coping found that coping had no discernable affect between attachment and distress. In contrast, Wei, Heppner, & Mallinckrodt (2003) examined perceived coping as a mediator between adult attachment and psychological distress in a college sample of over 500 students. Perceived coping fully mediated the relationship between attachment 25 anxiety and psychological distress and partially mediated the relationship between attachment avoidance and psychological distress. Their work demonstrated that attachment avoidance had both a direct and indirect effect on psychological distress, suggesting that other variables may be involved, such as different negative emotional states or behavioral confidence as suggested by Creasey & Hesson-McInnis (2001). Summarizing the current status of attachment research, there appear to be identifiable patterns of association between quality of attachment and coping behavior. For the most part, secure persons display predominantly adaptive coping in stressful situations, based upon their expectation that they can achieve fulfilling relationships with others, find ways to have their needs met, and evaluate distressing situations from a position of self-confidence and self-efficacy. People who are anxiously attached to significant others will predominantly operate from a position of rejection and fear, limiting their coping strategies to those of worry, hypervigilance, self-denigration, reactivity, and suppression. Those with avoidant attachment patterns operate from a position of mistrust, viewing their world as persistently unsafe. They find discomfort with intimacy and closeness and will strive to maintain physical and emotional distance from distressing cues, engage in denial, withdrawal, disengagement, and possibly substance abuse. Coping with Childhood Sexual Abuse Few studies have examined specific forms of coping with childhood sexual abuse and how these impact psychological functioning. Coffey et al. (1996) studied a community sample of 192 women who had been sexually abused as children and found 26 that disengagement was significantly related to general psychological distress even after controlling for abuse characteristics and dispositional styles of coping with other stressors. Disengagement was used more often to deal with CSA, while engagement strategies were used more often with other types of stressors. Lev-Wiesel (1999) studied a small sample of adults (24 men and 27 women, age 24 to 57 years) who experienced childhood abuse to determine the effect of potency (coping) on their feelings toward their offender-parents. Lev-Wiesel described potency as an ?enduring confidence in one?s own capabilities and confidence in? one?s social environment? that ?reflects the ability to maintain one?s emotional homeostasis in conditions where other resources lose their effectiveness (p. 294).? Fifteen percent of the sample reported being sexually abused during childhood, and potency was significantly negatively related to being sexually abused. Potency was also found to decrease the level of negative feelings toward the abuser. Since potency is described as both self-confidence and a belief in order and meaning in society, survivors of abuse who possess potency may have been better equipped to relieve hostility and revenge. They may have related to others with higher trust leading to more fulfilling interpersonal relationships. Merrill et al. (2003) studied the role of abuse severity, coping style, and sexual functioning in 547 female U.S. Navy recruits who reported childhood sexual abuse. They proposed that the severity of CSA affects the number of sex partners both directly and through its effect on coping and sexual functioning. Those women with greater CSA severity used more avoidant and self-destructive coping strategies than those with less severe CSA experiences. The use of self-destructive coping strategies was associated with dysfunctional sexual behavior and number of sex partners. Use of avoidant 27 strategies to deal with CSA was associated with higher number of sexual concerns and lower number of sex partners. While avoidant and self-destructive coping had opposite effect on number of sex partners, they were highly correlated with each other. This suggests abused women use both forms of maladaptive coping although these strategies may vary by situation or may vary over time. Sexual forms of coping with abuse. Many studies of childhood sexual abuse are confounded by the use of retrospective reporting, in that participant?s recall and reporting about traumatic experiences may be influenced by a number of confounds including their family environment, age and development, and processing of trauma. Hall et al. (1998) conducted one of few studies examining sexual behavior problems in children who were sexually abused. They found that children who were sexually abused displayed both self- focused sexualized behavior and interpersonal sexualized behavior as a result of several specific abuse characteristics. They found that sexual arousal of the child during his/her sexual abuse, the perpetrator?s use of sadism, and a history of emotional and physical abuse differentiated between those children with and without interpersonal sexual problems. Who the child blamed for the abuse (self or other) also differentiated between self-focused sexual behavior and interpersonal sexual behavior. The limited research into coping styles used by sexually abused males has focused primarily on sexualized forms of coping. Duncan and Williams (1998) studied adult men sexually abused as children and explored the differences in adult behavior as related to the gender of the offender and the use of force or coercion during the abusive act. Those men coerced as boys by older females were more likely to masturbate compulsively as 28 teens and to be sex offenders as adults. Survivors of such abuse also had higher levels of violence in adult intimate relationships than comparison groups of males involved in abuse by males or non-coercive acts by either gender. An explanation of sexualized coping provided by Marshall and Marshall (2000) states ?sex may be sought to achieve feelings of intimacy or to obtain affection, to alleviate boredom or a sense of frustration (non-sexual), as a way to obtain self- affirmation, to achieve a sense of conquest, or?to escape from problems (p.256).? Consistent with this view, they found that sexual offenders? juvenile histories include high relative rates of masturbation as a preferred way of coping with stress. Smallbone and McCabe (2003) also established a tentative link between attachment, early onset of masturbation, and sexual offending. They found that those offenders who reported CSA reported onset of masturbation 2.4 years ahead of those who were not abused (11 yrs vs. 13.4 yrs, respectively) and that onset occurred within a year of the abuse. Smallbone and McCabe proposed that early masturbation combined with deviant sexual fantasies creates classically conditioned deviant sexual proclivities that are acted upon in later adolescence and/or adulthood. Cortoni and Marshall?s (2001) Coping Using Sex Inventory (CUSI) was used to investigate sexualized coping within an incarcerated population. Sexual offenders, when compared to nonsexual violent offenders, were more sexually preoccupied during adolescence, and this preoccupation was related to later sexual offending. Intimacy deficits and emotional loneliness were predictive of greater use of sexualized coping strategies. Cortoni and Marshall?s inventory successfully distinguished between sexual offenders and nonsexual offenders with 74.3% accuracy, and demonstrated an internal 29 consistency of ? = .83. While Cortoni and Marshall proposed, ?in unskilled people, sex may easily come to serve as a coping mechanism to deal with loneliness and intimacy problems (p. 38),? they also recognized the need to determine if this phenomenon is unique to incarcerated men. Furthermore, they called for research into the frequency of using sex as a coping strategy in non-offender populations, along with obtaining the juvenile sexual history of non-offenders. In the case of men who have childhood sexual experiences, it may be important to understand masturbatory patterns (and associated fantasies) in order to determine if masturbation has a mediating effect among the general population of boys. Lyle (2003) followed this line of research by examining rates of sexualized coping in a non-offender sample and found over 55% of the college-age men reported masturbating every day or nearly every day for a period of a month or more. However, less than 16 % acknowledged being under stress at the time. Using the CUSI in the same sample, the majority of men who endorsed using sexual behavior when under stress identified consensual sexual behavior as their primary coping strategy. A smaller but significant proportion of the sample endorsed fantasizing and masturbating to themes of forced adult sex as well as consensual sex. There was a significant relationship between frequent masturbation and CUSI total score (r = .42, p<.01) and CUSI sex with adults score (r = .40, p<.01). Both sex with adults and fantasies about forced adult sex were highly correlated with the total CUSI score while items reflecting sex (or fantasies about sex) with children were not significant in this sample. Research has clearly identified sexualized forms of coping as short-term and long- term effects of abuse for men. The findings of Hall et al. (1998) suggest that sexual 30 arousal during abuse interacts with a child?s developmental maturity; the degree of pain, fear, and discomfort; and degree of self-blame to produce increasing levels of sexual behavior ranging from (a) expected sexual behavior, to (b) sexualized self-focused behavior, to (c) interpersonal sexual behavior. Arousal during the sexual abuse experience is related to a degree of self-blame for the abuse, and may be related in turn to developmentally premature introduction of masturbation. If the masturbatory practice becomes compulsive, as Marshall and Marshall (2000) suggest, and is associated with deviant fantasies, it follows that the sexual behavior might tend to be other-focused or interpersonal as reported in Hall et al. (1998). If, however, the compulsive masturbation is accompanied by non-deviant sexual fantasies as indicated by Lyle (2003), the use of sex may serve as a somewhat functional coping strategy or an emotional attachment alternative, thereby providing a measure of self-soothing. The suggested relationship between sexual arousal during abuse and later sexualized behavior appears to mesh nicely with models that include either attachment disruption or sexualized coping as mediators in the potential outcome(s) of childhood sexual experiences. However, it is unclear at this juncture whether the presence of sexualized coping is sufficient to explain offending behavior, since non-deviant sexualized coping is common in non-offenders. Rather, it may be the restricted range of other coping behaviors, particularly problem-focused behavior, which creates the dependency on sexual means for relieving distress. Child Sexual Abuse, Attachment, and Coping Current techniques in structural equation modeling make it possible to evaluate more complex relationships between several variables. It is now possible to develop 31 models to examine how childhood sexual experiences, attachment to parents and friends, and use of specific coping strategies may interrelate to produce long-term psychosocial effects (Marshall & Marshall, 2000; Marshall, Serran, & Cortoni, 2000; Merrill, Thomsen, Sinclair, Gold, & Milner, 2001; Merrill, Guimond, Thomsen, & Milner, 2003; Roche, Runtz, & Hunter, 1999). The literature provides ample support for examining these variables in relation to both offending and non-offending behavior. Such a model should identify childhood sexual experiences that may be abusive, assess adult attachment patterns, and evaluate dispositional and situational coping behaviors victims later use when distressed. Marshall and Marshall?s (2000) sexual-offending model suggests a child who is insecurely attached is at greater risk for vulnerability to abuse, and possesses fewer emotional resources for coping with his abuse. When a boy is introduced to sexual activity he must then incorporate the degree of fit this experience has with his emotional development, his preconceived notions of male-ness, and his awareness of his sexual identity (Gartner, 1999). The incongruity between his gender- and sexual identity and his current sexual experiences may be compounded by other factors, such as sexual coercion by authority figures and trusted individuals, as these situations also may violate the boy in terms of developmental ability to recognize his betrayal and lack of power (Finkelhor, 1985). For those boys who have childhood sexual experiences incongruent with their development or sexual identity, it can be assumed that some emotional injury may occur that necessitates the use of coping strategies. Once again, these coping strategies are governed by the social and emotional resources available to the boy, and his level of 32 maturity in dealing with adverse circumstances. Sexualized coping strategies, including compulsive masturbation and sexual coercion, have been proposed as the preferred coping method used by sex offenders to relieve distress. According to Marshall & Marshall?s model (2000), for those boys who use sexualized coping, it is assumed that these behaviors and fantasies become associated through behavioral conditioning, probably through reinforcement of the boy?s fantasized sexual object choice. This conditioning process sets the stage for deviant arousal patterns. However, in recent work examining sexualized coping in non-offenders (Lyle, 2003), sexualized coping was frequently used within the sample, focused primarily on non-deviant themes, and was significantly related to childhood sexual experiences. This finding suggests that more needs to be learned about the coping strategies of offenders and non-offenders. For example, differences between attachment to parents and peers or between mother and father may produce differential effects on coping style. In Lyle?s 2003 pilot study, attachment anxiety was related to frequent masturbation. Difficulties with parent attachment predicted increased sexualized coping in general, difficulties with peer alienation produced a tendency toward sexual coercion, and difficulties with parent alienation produced a tendency toward fantasizing about sex with children. Mother?s care was negatively related to forced adult sex and father?s care was negatively related to sex with children. These preliminary findings in a relatively small sample (107 college men) coincide with models of offender development, but fail to explain why these men were not offending. The present research will attempt to explore a broader range of coping behaviors given that sexualized coping was endorsed by many non-offending college men. 33 Attachment and coping research suggest that men who have had early childhood sexual experiences that are abusive will respond to those experiences in ways consistent with their attachment to their caregivers. Maternal and paternal attachment may have some independent effect on the coping behaviors used by CSA victims. Anxious and avoidant attachment also should produce different coping strategies, compared to securely attached men who will rely primarily on problem-focused strategies. Ultimately, the dispositional and situational coping styles of men should predict their overall psychological functioning and also may predict specific psychological difficulties. Hypotheses (1) Attachment patterns will predict specific coping strategies in men who have experienced childhood sexual abuse. Anxiously attached men who have had childhood sexual experiences, particularly those they can identify as abusive, will use sexualized coping when under stress, as found by Lyle (2003). According to Alexander (1992), these men might be best described as having a preoccupied attachment, and will therefore engage in increased worry and more desperate attempts to merge with others. These individuals will predominantly use frequent masturbation and consensual interpersonal sex when under stress. Creasey & Hesson-McInnis (2000) suggest anxious individuals demonstrate decreased emotional regulation and increased conflict negativity. Therefore, anxious men may also engage in relatively more emotion-focused strategies than their peers. Men with avoidant attachment styles who have had childhood sexual experiences will rely less often than their anxious peers on masturbation and consensual sex when 34 under stress (Lyle, 2003). However, because avoidant persons experience isolation, dependency, and lack of trust (Alexander, 1992) they are also likely to use fewer adaptive coping responses (Creasey & Hennis-McInnis, 2000) and rely more often on disengagement and denial when under stress (Mikulincer, 1998). In line with basic attachment models (Berk, 1998; Bowlby, 1988; Thomas, 2000), securely attached men who have had childhood sexual experiences will have greater skills related to intimacy, and will possess internal working models of the self as competent. Therefore, they will rely more often on problem-focused strategies, support- seeking, and reframing than their peers. The differential effects of attachment to mother, father, and peers also may predict different sexualized coping preferences in a sample of non-offenders. Extending previous findings by Lyle (2003), poor maternal attachment will predict fantasies and behavior related to forced sex with adults; poor paternal attachment will relate to sexual fantasies and behavior directed toward children; and poor attachment to peers will predict more frequent use of sexual coercion. (2) Insecure men who recognize their childhood sexual experiences as abusive will have more problematic dispositional coping strategies, including disengagement and sexualized coping, than their securely attached peers. Abuse severity has been implicated in the long-term effects of CSA. Merrill et al. (2003) found in a female sample that abuse severity was related to increased self- destructive coping, higher numbers of sexual partners, and sexual dysfunction. Thomas (2000) proposed that severe abuse results in an internal working model of the endangered child, inadequate protector, and uncontrolled aggressor. These work together to produce a 35 negative internal dialogue that limits coping behaviors and produces depression, self- injury, and victimization of others. Individuals who have failed to achieve mastery of certain stressful experiences will likely possess impaired self-efficacy and inadequate coping resources, due in part to severe abuse and poor caregiver models (Bandura, 1997). Additionally, only those men with the most severe forms of abuse will identify themselves as victims of childhood sexual abuse (Stander et al., 2002). Therefore, those men who self-identify will report the most problematic dispositional coping. (3) If childhood sexual abuse is significantly related to both attachment and coping, and attachment exerts control in the use of coping strategies, then attachment should mediate the relationship between abuse and coping. Previous research has identified relationships between CSA, attachment, and sexualized coping (Lyle, 2003; Marshall & Marshall, 2000; Smallbone &McCabe, 2003), and mediational models have been supported. Therefore, men with the strongest bonds with parents and friends, and who have experienced childhood sexual abuse, will be the most likely to utilize adaptive coping in response to current stressors. Conversely, poorly attached men who have experienced childhood sexual abuse will have more problematic coping responses when distressed. (4) Coping strategies are expected to predict psychological adjustment, and therefore should mediate the relationship between CSA and psychological functioning. Coffey et al. (1996) found in a female sample that disengagement coping strategies were related to increased general psychological distress. Merrill et al. (2003) produced a model that indicated the effect of abuse severity on later functioning was partially mediated by coping strategies, with negative strategies producing the strongest 36 relationship to symptoms. Lyle (2003) also found moderate support for sexualized coping mediating the effect of CSA on psychological adjustment. It is therefore reasonable to investigate the role of both sexual and non-sexual coping in mediating the effect of CSA on psychological adjustment. Men who tend to rely upon problem-focused and support-seeking strategies will more often provide a psychological profile within the normal range, and will tend to report fewer stress-related psychological sequelae than their peers. Men who use more emotion-focused strategies will more often provide psychological profiles indicating difficulty with affect regulation and stress-related effects such as interpersonal problems and substance abuse. Men who use avoidant, sexualized, and denial strategies will likely provide more pathological psychological profiles and will report the most difficulty with interpersonal relationships, substance use, antisocial behavior, and affect regulation. Preliminary Models Figures 1 and 2 provide simple models of proposed relationships (adapted from Wei, Heppner, and Mallinckrodt, 2003). Figure 1 demonstrates that attachment anxiety and attachment avoidance will exert differential influence on coping behaviors and psychological adjustment. Similarly, the differential effect of attachment to mother and father on coping strategies and psychological adjustment is indicated in Fig. 2. 37 Mediational Model: CSA ? Attachment ? Coping ? Psychological Adjustment Fig 1. The Impact of Attachment Style Fig 2. The Impact of Attachment Figure Attachment Anxiety Attachment Avoidance Psychological Adjustment Coping Child Sexual Abuse Father Attachment Mother Attachment Psychological Adjustment Coping Child Sexual Abuse Peer Attachment 38 METHOD Participants Male college students were recruited from the psychology department and criminology department at a southeastern (home) university, and from other universities through an online PsyTeacher listserv. Participation in the study provided students extra credit points in their undergraduate psychology courses. Participants were advised that the research concerned family environment, sexual experiences, and current functioning. A Written Information Letter and psychological referral resources were provided to the students prior to their participation in the study. Students were advised that some of the requested information was of a sensitive nature, and that no identifying information would be collected. Students also were advised that they could discontinue the study at any point without penalty. Demographic information was collected including age, race/ethnicity, marital status, as well as gender and sexual orientation. Instruments A modified version of the Childhood Sexual Experiences Checklist (CSEC) developed by Stander et al. (2002) was used to examine childhood sexual experiences. The CSEC asked participants about sexual experiences they may have had with a family member prior to their 18 th birthday that involved sexual touching or sexual penetration. It also asked about sexual experiences before age 18 with non-family members who were at 39 least 5 years older. The CSEC allowed participants to identify the perpetrator, the relative age of the perpetrator and victim at the time of the encounter (5 years older or younger), and whether force or threats were used. The survey also collected demographic information. Modifications to the CSEC include eliminating some demographic questions irrelevant to this research and adding multiple questions about early childhood sexual experiences. Examples of such questions include ?At what age did you begin masturbating?,? ?At what age did you have your first sexual experience with another person?,? ?Have you ever been aroused by another person when you did not want to be?,? ?Did having an erection under these circumstances bring up uncomfortable emotions for you?,? and ?Have you ever experienced a period of time in which you masturbated every day or nearly every day for a month?? Frequency and situational distress were also measured. Attachment style was measured using the Inventory of Parent and Peer Attachment (IPPA; Armsden & Greenburg, 1987) and the Experiences in Close Relationships Inventory (ECL; Brennan, Clark, & Shaver, 1998). The IPPA asked respondents to rate 53 statements regarding current attachment functioning with parents and friends on a 5-point Likert scale, producing scales for mother, father, and peer attachment and subscales related to communication, trust, and alienation. The ECL asked respondents to rate 36 statements on a 5-point Likert scale reflecting two main factors of anxious or avoidant attachment. ECL scores also were interpreted in terms of secure, dismissive, fearful, and preoccupied attachment. Cortoni and Marshall?s (2001) Coping Using Sex Inventory (CUSI) was used to measure the degree of sexualized coping in the sample of college men. The CUSI 40 consists of 16 statements reflecting ways people react to various difficult, stressful, or upsetting situations, rated on a 5-point Likert scale. Sexualized coping strategies identified by the CUSI reflect 3 factors including consensual adult themes, rape themes, and child sexual abuse themes. Carver?s Brief COPE (1997) was used to measure coping behaviors in the sample. The Brief COPE consists of 28 items obtained from research using the full version of the COPE (Carver et al., 1989) and yields 14 coping styles (2 items each). The Brief COPE omits Restraint Coping and Suppression of Competing Activities scales from the full measure; refocuses Positive Reframing, Venting, and Self-distraction; and adds a scale for Self-blame. A 5-point Likert scale was used to rate participants? use of each strategy from I haven?t been doing this at all to I?ve been doing this a lot. The Brief COPE was presented in a dispositional format and current situational format. The Personality Assessment Screener (PAS; Morey, 1997) was used to measure participants? psychological adjustment. The PAS consists of 22 items across 10 scales and is designed to provide a brief assessment of clinically significant difficulties. The test items were selected by factor analysis from the Personality Assessment Inventory for their ability to predict clinically significant scores (Morey). All survey instruments can be found at Appendix A. Procedure Males age 19 or older were invited to participate in this research project. Because data collection was completely anonymous, informed participation methods were used rather than informed consent. All participants were advised that the research was concerned with family environment, sexual experiences, and current functioning, and that 41 the questions may ask for information they consider private and of a sensitive nature. Each participant received a questionnaire containing the ECL, IPPA, CSEC, CUSI, Brief COPE, and PAS. Completion time for each packet was expected to be approximately 45 minutes. Packets were administered in three forms: a paper form of the questionnaire was administered to approximately 25 participants; an electronic format of the questionnaire was provided at computer terminals for approximately 50 participants; and a web-based form of the questionnaire was made available to approximately 200 participants. In each case, an extra credit voucher was provided to students for participation if they so desired. Electronic Internet-based research. Particular issues arise in conducting Internet-based surveys and experiments. Advantages of Internet research include access to a widely dispersed and diverse pool of research participants, much faster accumulation of data, and ease of recording directly into databases (Benson, 2003). Challenges include sample bias, participant identity verification, and controlling experimental conditions. Ethical challenges and considerations are also of paramount concern, and steps were taken to ensure that ethical considerations currently applied to traditional research were adapted to fit the demands of Internet research (Frankel & Siang, 1999). Birnbaum (2000) addresses general questions pertaining to the agreement between the research findings of Internet and laboratory experiments and reports no significant differences in findings. Demographics and motivations of people who participate in Internet research, information about the first Web studies, and methodological considerations for Internet research are also reviewed by Birnbaum. 42 One important aspect of Web experimenting is that the experiment comes to the participant rather than the reverse (Reips, 2000). Web experiments spare both researchers and participants the difficulties of scheduling time and space, among other benefits. Soon researchers may rely heavily on the point-and-click behavior of Internet users to attract, coordinate, and record participants and their relevant data. Currently, web-based psychological research methods are appearing at an increasing rate, with methodological and technical support evolving steadily (Birnbaum, 2000). One of the most relevant ethical concerns with this research project was the protection of participant anonymity. With anonymity comes the risk that participants may use deception in their responses. Participants who accessed the web-based survey were required to provide information about gender and birth date before entering the protocol. At that point, those who met screening criteria (males age 19 years or older) were assigned a participant number and required to create a Personal Identification Number (PIN). At that time, they received specific information about the nature of the survey and other material pertinent to informed participation. The researcher?s email and contact information were made available to participants who wished to communicate one on one, and anonymity was maintained as participant number and participant email or Internet Protocol (IP) address were not associated. Obtaining informed consent provides some challenge in many Internet research projects (Frankel & Siang, 1999), however it is not a large concern in cases with informed participation procedures. Participants who entered this study electronically were provided a web page containing information adapted from the paper Informed 43 Participation statement. A web script was devised requiring a response on an Accept or Decline button to proceed or leave the survey. Data collection was accomplished via Javascripts provided by University of California Fullerton at http://psych.fullerton.edu. The university makes the scripts available at no charge and provides a data storage service as well. Numerous research- oriented sites are available for producing, advertising, and conducting web-based surveys as well (Birnbaum, 2001). Survey instruments were designed using Birnbaum?s scripts, and provided to the home university?s College of Liberal Arts computer web design office for modifications. Data storage was accomplished by utilizing secure web-space provided by the home university. The current project was designed for multiple methods of data collection and no single format was intended to predominate. Paper and pencil methods of data collection present the most difficult challenges for time demands, space availability, participant pools, and data recording accuracy. Electronic data collection at research sites (university settings) can provide a more streamlined method for managing large quantities of data, but is somewhat limited again by time requirements, space requirements, and multiple IRB reviews. Internet data collection presents more technical challenges in web design, advertisement and recruitment, and compensation (extra credit, for example). However, the benefits of larger sampling and automated data collection made this an appealing methodology. This project was designed to stand upon any or all survey methods. Analyses The current project expanded upon previous work linking attachment and sexualized coping, and mediational models relating the effects of attachment and coping 44 on psychological outcome (Lyle, 2003). Variables of interest paralleled those examined by Stander et al. (2002), Cortoni & Marshall (2001), and Roche et al. (1999). Independent variables included demographic information, various childhood sexual experiences, researcher-defined versus self-defined CSA, developmental/gender/sexual identity congruence, and childhood attachment. Mediating variables included the degree of adult attachment to mother, father, and peers, participants? use of compulsive masturbation and/or 3 sexualized coping strategies provided by the CUSI, and the coping styles revealed by the Brief COPE. Outcome variables include psychological adjustment data provided by the PAS, comparing total scores and subscale scores. Preliminary analysis first established individual relationships between the four main variables of interest, CSA, attachment, coping, and psychological adjustment. Then potential mediating effects were tested. According to Baron and Kenny (1986), tests of mediation require that ?(a) variations in the predictor variable significantly account for variations in the presumed mediator, (b) variations in the mediator significantly account for variations in the independent variable, and (c) when paths a and b are controlled, a previously significant relation between the independent and dependent variables is no longer significant?? Baron and Kenny also assert that multiple mediators exist when the significance of path c is not reduced to zero by controlling for paths a and b. Figures 1 and 2 provide initial models for examining the relationships between latent variables within the domain of childhood sexual abuse. Attachment, whether examined in terms of anxiety and avoidance or comparing mother to father, is expected to exert some control in the effect of abuse on coping and psychological adjustment. When coping behavior is 45 examined at the subscale level, it is expected that poorer attachment will relate to specific maladaptive coping behaviors and thus relate to poorer psychological adjustment. Hypothesis 1: Attachment patterns will predict specific coping strategies in men who have experienced childhood sexual abuse. As with Cortoni & Marshall (2001), this research attempted to determine participants? use of sexualized, disengaging, or emotional coping strategies to deal with loneliness, intimacy problems, and childhood/juvenile sexual experiences. By using the CUSI, 3 main factors were identified among the sexual strategies used by college men, including fantasized consensual sexual behavior, rape themes, and child sexual abuse themes. While Cortoni & Marshall used the CUSI to predict sexual offender status, Lyle (2003) found that similar sexualized coping strategies appeared in non-offender populations. In addition to the CUSI, the Brief COPE (Carver et al, 1989) provided dispositional and situational profiles indicating various coping strategies for dealing with stress. It was also expected that Carver et al.?s scales could delineate coping styles according to the five-factor solution provided by Finch et al. (1999). Testing hypothesis one, participants? attachment scores on the ECL and IPPA were correlated with two indicators of sexualized coping behavior. The first indicator of sexualized coping was participants? response to the question ?Have you ever experienced a period of time in which you masturbated every day or nearly every day for a month?? adapted from Duncan and Williams (1998). A follow-up question asked if the period of masturbation was related to a stressful event or time in the participant?s life. The second indicator of sexualized coping was the total score provided by Cortoni and Marshall?s 46 (2001) Coping Using Sex Inventory (CUSI). Similarly, attachment scores were correlated to the Brief-COPE?s situational and dispositional coping factors. The attachment scores from each measure provided varying ways to represent attachment (Anxious/Avoidant with the ECL; and Communication/Trust/Alienation with the IPPA). Therefore, each subscale score was examined independently and across scales to evaluate the use of problem-focused, emotion-focused, disengagement, and sexualized coping strategies within each domain. Additional analyses also were possible along the three identified CUSI factors: sex with consenting adults; forced sex with adults; and forced sex with children. Once the significant relationships between attachment and coping were identified, the relevant variables were used to create a correlation matrix, which was then used to develop an exploratory model (using Structural Equation Modeling with AMOS 4.01, SmallWaters Corporation, 1994-1999) linking attachment and specific coping behaviors. Confirmatory factor analysis was then used to refine the model, specifically using the AMOS 4 Regression Weights (with corresponding p-values) and the Modification Index to systematically eliminate variables and link error terms in an effort to improve the model?s fit. Fit measures used to support the model include the AMOS 4 Minimum Discrepancy indicator (? 2 /degrees of freedom), Goodness of Fit Index (GFI), the Normed Fit Index (NFI), and the Root Mean Square Error of Approximation (RMSEA) with its p- value for test of close fit (cut point set at .05). A ? 2 /df ratio between 2 and 3 was considered to be a good fit; a GFI and NFI >/ = .9 was set as the value for a good fit; and RMSEA value below .05 was considered a very good fit (with p-values above .05). 47 Hypothesis 2: Insecure men who recognize their childhood sexual experiences as abusive will have more problematic dispositional coping strategies, including disengagement and sexualized coping, than their securely attached peers. As in the pilot research (Lyle, 2003), childhood sexual experiences were correlated with several dependent variables to examine significant relationships. Among these dependent measures are variables suggesting family violence, types of adult sexual experiences, use of mental health services, and indicators of emotional distress (Stander et al., 2002). Abuse severity included information about number of occurrences, relationship to the abuser, and use of force. Participants? attachment scores and their specific attachment to mother, father, and/or peers were used to classify them into attachment groups. Multiple regressions were used to predict coping factors based on attachment group classification and severity of abuse. Hypothesis 3: If childhood sexual abuse is significantly related to attachment, and attachment exerts control in the use of coping strategies, then attachment should mediate the relationship between abuse and coping. Participants were grouped by their exposure to various forms of childhood sexual experiences (CSEs). Group 1 consisted of those participants who defined themselves as victims of childhood sexual abuse, Group 2 included men who reported CSEs with any immediate or extended family member, and Group 3 included men who experienced CSEs, either by consent or force, prior to age 13 with any person at least 5 years older. Group 4 consisted of men who reported sexual contact with a child 5 or more years younger, and Group 5 included men who did not report childhood sexual experiences. Groups were compared in their use of compulsive masturbation (as assessed by the 48 question above), their overall CUSI score, and their coping scores. An aggregated researcher-defined CSA group also was created and coping was analyzed at this level as well. CSA (by group and as a whole) was evaluated as a predictor of attachment, using the significant attachment subscales identified in hypothesis one. Subsequently, the CSA ? Attachment ? Sexualized coping model was created and tested using identical procedures with AMOS 4.01 (as indicated for Hypothesis 1). Hypothesis 4: Coping strategies are expected to predict psychological adjustment, and therefore should mediate the relationship between CSA and psychological functioning. Additional analyses tested the mediating effect of coping on adjustment by determining if CSA predicted both coping and psychological adjustment, and if coping (testing compulsive masturbation and coping factors independently) predicted psychological adjustment. These analyses incorporated the ideas of a number of researchers and provided the basis for beginning to dismantle the non-offending effects of CSA on men. To test hypothesis 4, participants were grouped by CSA (as above in Hypothesis 3). Sexualized coping was determined by the dichotomous-choice question regarding compulsive masturbation and scores on the CUSI. Dispositional coping strategies were determined by the Brief COPE. Outcome measures were represented by overall score on the PAS, and were evaluated at the total score and the subscale level. Once significant variables were identified (as in earlier hypotheses), SEM procedures were again used to create and test a model representing the correlations 49 between CSA, coping, and psychological adjustment. Mediation was determined by testing the change in significance created by entering coping variables into the model. 50 RESULTS Overview One hundred sixty-one participants completed all of the research instruments during one semester at a Southeastern university. Of these, 22 packets were completed in paper form and 33 were completed on a computer terminal in the university?s psychology lab. Lab participants were randomly assigned to the paper or computer administration using a random number list and assigning odd numbers to paper and even numbers to computer. The remaining 106 completed surveys were completed online by men who responded to announcements to other university departments, announcements on PsyTeacher listserv, or to a web search for psychology research. There were 222 total entries in the online database, with over 50% dropping out before completion. The online survey required completion of six separate instruments to be included in analysis, and these instruments matched those provided on paper and on computer terminals. Survey Format Participants indicated the format/origin of their survey with 0- Paper (in the university lab), 1- Computer (also in the lab), 2- Other university department, 3- PsyTeacher listserv, 4- American Psychological Association (APA) listserv, and 9- Other (or no response). A discrepancy occurred in the reported number of lab participants (51 versus 33) that must be attributed to online participants indicating the wrong format. Frequencies are provided in Table 1 and Figure 3 below. 51 Table 1. Survey Format Format Frequency Valid Percent Valid 0 - Paper 22 13.7 1 - Lab Computer 51 31.7 2 - Other Department (online) 29 18.0 3 - PsyTeacher listserv (online) 48 29.8 4 - APA listserv (online) 2 1.2 9 - Other (online) 9 5.6 Total 161 100.0 Figure 3. Survey Format 5.6% 1.2% 29.8% 18.0% 31.7% 13.7% 9- Other 4- APA listserv 3- Teacher Listserv 2-Other Dpt (online) 1- Computer Lab 0- Paper One-way analysis of variance was used to examine significant differences between groups. Results indicated significant differences between groups for attachment total scores (mother, father, and peers), age, education, parental violence in the home, suicidal ideation and attempts, situational cognitive reframing (coping), and dispositional problem-solving. Table 2 summarizes the results. 52 Table 2. Differences by Survey Format Sum of Squares df Mean Square F Sig. MOTHER Between Groups 13666.567 5 2733.313 7.583 .000 Within Groups 55868.427 155 360.441 Total 69534.994 160 FATHER Between Groups 14373.240 5 2874.648 4.814 .000 Within Groups 92565.257 155 597.195 Total 106938.497 160 PEERS Between Groups 6693.397 5 1338.679 5.959 .000 Within Groups 34819.709 155 224.643 Total 41513.106 160 AGE Between Groups 2164.963 5 432.993 13.007 .000 Within Groups 5159.820 155 33.289 Total 7324.783 160 EDUCATION Between Groups 11.029 5 2.206 4.351 .001 Within Groups 78.574 155 .507 Total 89.602 160 F_VIOLEN Between Groups 22.481 5 4.496 2.753 .021 Within Groups 253.158 155 1.633 Total 275.640 160 M_VIOLEN Between Groups 19.532 5 3.906 3.723 .003 Within Groups 162.642 155 1.049 Total 182.174 160 SUI_ID Between Groups 3.432 5 .686 3.416 .006 Within Groups 30.943 154 .201 Total 34.375 159 SUI_ATT Between Groups .873 5 .175 3.530 .005 Within Groups 7.620 154 .049 Total 8.494 159 S5REFRAM Between Groups 708.006 5 141.601 4.755 .000 Within Groups 4615.770 155 29.779 Total 5323.776 160 D5PROBLM Between Groups 684.574 5 136.915 2.972 .014 Within Groups 7141.202 155 46.072 Total 7825.776 160 Note. F_VIOLEN = Father violence; M_VIOLEN = Mother violence; SUI_ID = Suicidal ideation; SUI_ATT = Suicide attempts; S5REFRAM = Situation reframing; D5PROBLM = Dispositional problem-focused coping. Post hoc comparison of significant mean differences indicated that those participants who completed paper or computer packets in the university lab (with the investigator present) provided higher maternal attachment scores than those in the other 53 four collection formats. For paternal attachment, those in the paper condition (Format 0) provided higher scores than two of the online conditions (PsyTeacher listserv ? 3, and Other ? 9). The two lab conditions also provided significantly higher peer attachment scores than the other online conditions. Formats 0 ? 3 included college participants whose age was significantly lower, and level of education was less, than formats 4 and 9 (APA listserv and Other). Online psychology students (Format 3) reported significantly more paternal violence than those students in the paper condition, but were similar to all others. The college students (Formats 0 ? 3) reported lower maternal violence scores than the two older groups (Format 4 and 9). Format 9 participants reported significantly higher suicidal ideation and attempts than the other participant groups. The two lab conditions provided higher situational reframing scores than the online psychology students, and higher dispositional problem-solving scores than the APA listserv participants. Differences in age and group size may be plausible explanations for these mean differences in scores. However, it also may be that participants in the lab condition, with the female investigator present, provided more socially desirable responses relating to attachment and family violence. Conversely, one might argue that those participants who originated from online coursework and internet surfing are somewhat socially isolated and less reliant upon others. They also may be less inhibited in a testing situation that might be more private. After examining the differences by format, it seemed necessary to limit some of the variability by restricting the age of the sample. However, some of this variability had to be maintained to produce adequate models of abuse ? effect relationships. Procedures for these age restrictions are described later. 54 Demographic Characteristics The mean participant age was 22.09 (SD = 6.76) ranging from 19 to 67. Over 91% of the sample was younger than 27 years old while 13 participants (8.1% of the sample) were between age 27 and 67. The racial makeup included 87% Caucasian, 6.2% African American, 3.7% Asian American, 1.2% Hispanic American, and 1.9% Other Race/ethnicity. Participants? education varied from high school to a masters or higher degree with 1.9% reporting some high school, 17.4% high school diploma or GED, 75.8% reporting some college, 1.2% with a bachelors degree, .6% in graduate school, and 3.1% with a masters degree or higher. Marital status included 91.3% single, 5% married, 1.9% divorced, and 1.9% cohabitating. Ninety-two percent of the sample had no children while 2.5% had one child, 2.5% had two children, and one participant had three children. Seventy-one percent of the sample reported their parents were still married while 20.5% reported their parents were divorced. Of those whose parents divorced, the mean age of the respondent at the time of divorce was 8 years old (SD = 6.87). Family environment. Quality of home environment is considered to contribute to abuse (Merrill et al., 2001; Rind et al., 1998; Alexander, 1992); therefore participants were asked how involved their parents were in raising them, indicated on a 5-point Likert scale ranging from 1- not at all involved to 5- extremely involved. The mean score for mother involvement was 4.66 (SD = .72) and the mean score for father involvement was 4.01 (SD = 1.13). The men rated how strict their parents were on a 5-point scale from 1- not at all strict to 5- extremely strict with a mean score of 3.61 (SD = 1.0). Fifty-eight percent of the men reported no violence in the home committed by the father and 69.6% reported 55 no violence committed by mother. For those who experienced violence in the home, rates of violence committed by the father included 26.7% reporting 1-5 incidents and 15.5% reporting 6- 20+ incidents. Similarly for mother violence, 21.1% reported 1-5 incidents while 9.3% reported 6-20+ incidents. Mental health. Mental health issues have been identified as correlates of abuse (Fromuth & Burkhart, 1989; Smallbone & Whortley, 2000; Lyle, 2003) and are therefore reported here. Suicidal ideation was reported by 31% of the sample, while 5.6% reported a suicide attempt. Seventy-two percent of the sample had no contact with the mental health system, while 10.6% reported 10 or more visits with a mental health professional. More than 14% reported having 1 to 8 visits with a mental health counselor. Sexual History Participants? sexual history included questions regarding age of onset for masturbation and interpersonal sexual behavior, issues related to early sexual orientation and acts that may have run counter to their sexual orientation, types of contact that may have included force or coercion, unwanted sexual arousal, compulsive masturbation, weekly/monthly masturbation rates, number and gender of sexual partners, and use of condoms during sex. Sexual orientation within the sample included 96.3% heterosexual, 1.2% homosexual, 1.2% bisexual, and 1.2% unsure of their sexual orientation. Thirteen percent of the sample reported experiencing a sexual act contrary to their sexual orientation. Forty-two percent of the sample reported being sexually aroused by another person when they did not want to be. 56 Since masturbation is considered a key variable in understanding sexualized coping, this behavior was measured in several ways. First, men were asked ?Have you ever experienced a period of time in which you masturbated every day or nearly every day for a month?? Among all participants, 59.6% responded affirmatively (termed Compulsive Masturbation: Comp. Mast). Additionally, 100 participants responded to the question ?Was this period of masturbation related to a stressful time in your life?? with 21% of those men (13% of the sample) responding affirmatively (termed Masturbation under Stress: Mast_Stress). In addition to these dichotomous choice variables, 7-day and 30-day masturbation rates were collected for the period immediately preceding the completion of the survey. One hundred fifty-nine participants (n = 159) provided 7-day rates with a mean of 2.84 (SD = 3.11) including 20.8% reporting none, 72.9% reporting 1 to 7 times, and 6.3% reporting 8 to 24 times in the previous 7-day period. One hundred fifty-five men (n = 155) provided 30-day masturbation rates with a mean of 12.08 (SD = 11.15) including 11.6% reporting none, 62.6% reporting 1 to 15 times, 20.6% reporting 16 to 30 times, and 5.2% reporting 31 to 67 times in the previous 30 days. Summary information from the sexual history is provided in Table 3, indicating means, standard deviations, frequencies, and percentages for several pertinent sexual variables. The columns represent figures for the full sample, an age-restricted subset (mean age 21), and a subset of men who responded to questions related to compulsive masturbation. The Comp. Mast. subset indicated higher current rates of masturbation, and higher rates of forced sex and sexual abuse before age 13. 57 Table 3. Sexual History Variables Full Sample (n = 161) Age-restricted Group (n = 151) Comp. Mast Group (n = 100) Mean (SD) Mean (SD) Mean (SD) Onset of masturbation 12.56 (3.57) 12.58 (3.64) 12.94 (2.55) First sexual experience 13.11 (5.67) 13.07 (5.59) 13.02 (5.36) 7-day masturbation rate 2.84 (3.11) 2.93 (3.17) 3.72 (3.50) 30-day masturbation rate 12.08 (10.0) 12.41 (11.26) 15.83 (12.10) Sexual orientation Frequency (%) Frequency (%) Frequency (%) Heterosexual 155 (96.3%) 145 (96%) 95 (95%) Homosexual 2 (1.2%) 2 (1.3%) 2 (2%) Bisexual 2 (1.2%) 2 (1.3%) 2 (2%) Unsure 2 (1.2%) 2 (1.3%) 2 (2%) Sexual dissonance (yes) 21 (13%) 16 (10.6%) 15 (15%) Unwanted arousal (yes) 68 (42.2%) 60 (40.4%) 42 (42%) Compulsive Masturbation (yes) 96 (59.6%) 90 (59.6%) 94 (94%) Masturbation under stress (yes) 21 (13%) 17 (11.3%) 21 (21%) Force 13 (yes) 7 (4.3%) 6 (4.0%) 6 (6.0%) Force 18 (yes) 3 (1.9%) 3 (2.0%) 2 (2.0%) CSA 13 (yes) 6 (3.7%) 4 (2.6%) 5 (5.0%) CSA18 (yes) 3 (1.9%) 3 (2.0%) 3 (3.0%) Note. Sexual dissonance = experience contrary to sexual orientation; Force 13 = forced sex before age 13; Force 18 = forced sex age 13-18; CSA 13 = ?child sexual abuse? before age 13; CSA 18 = ?child sexual abuse? from age 13-18. The Coping Using Sex Inventory (CUSI) measured sexualized coping including masturbatory and interpersonal sexual behavior and fantasies. The sample mean for the CUSI was 28.48 (SD = 6.71) with scores ranging from 16 (minimum possible score) to 50. Reliability estimate for the CUSI was ? = .89. Participants also were asked to report pressuring or forcing someone to have sex, and to report how many times they had done so. Nineteen men (11.8% of the sample) reported pressuring someone for sex. The average number of times pressuring occurred was 3.35 (SD = 6.45) times, ranging from 0 to 30 times. Only 27 men responded to questions about using force for sex, and among them 22.2% (3.7% of the sample) admitted forcing someone to have sex. The mean score for the number of times this occurred was 5.56 (SD = 9.71), ranging from 0 to 25 times. Current sexual behavior also included number of partners and use of condoms. The mean number of female sexual partners was 6.64 (SD = 10.8) ranging from 0 to 100, 58 and for male partners the mean was .2 (SD = .78) ranging from 0 to 8. Of 156 men responding, 22.4% never use condoms, 27.6% use them sometimes, 22.4% use them almost always, and 27.6% use them always. Childhood sexual experiences. When asked about being forced into sex of any kind, 4.3% of the sample reported forced sex before age 13 and 1.9% reported forced sex from age 13 to 18. When the question was stated as being sexually abused as a child, 3.7% reported CSA before age 13 and 1.9% reported CSA between age 13 and 18. Further questions served to distinguish family sex from non-family sex, penetrative acts from touching and kissing, and acts committed by someone 5 or more years older or against someone 5 or more years younger from those involving age-mates. Additionally, participants were asked to rate how upsetting these experiences were, at the time and currently, on a 5-point scale from 1- not at all upset to 5- extremely upset. Ten percent (10%) of those responding to family sex questions (n = 154) reported sexual acts with family members, including 15 men (9.3%) experiencing sexual touching and 4 men (2.5%) reporting penetrative acts. Among 17 men reporting family sex, 47% had sexual contact with a sibling, 35% with a cousin, 6% with a grandparent, and 6% with an aunt or uncle. Forty-one men responded to questions regarding family sex with someone older with 3.7% of the sample (14.6 of the set) reporting acts with someone 5 years older. One individual admitted sexual contact with a family member 5 or more years younger. Rates for non-family sex were much higher, as these acts included typical dating behaviors. Abusive acts were distinguished by force or coercion, a significant age difference, acts by a person of authority or trust, or acts that resulted in some degree of 59 distress. Nearly 56% reported non-family sexual acts before the age of 16. This lower age was chosen as it relates closely to the age of legal consent. Among 64 men who indicated the relationship of the sexual partner, 81.3% were boyfriend or girlfriend, 12.5% were an acquaintance or stranger, 4.7% indicated sex with a babysitter, and 1.6% indicated sex with a teacher or coach. Among 97 men responding to specific non-family acts, 36% reported non-family penetrative acts while 2.1% experienced forced acts with non-family members. Among 99 respondents, 10% reported sexual acts with non-family members 5 or more years older. Three men (2% of the sample) admitted having sex with non-family members 5 or more years younger. Non-family sex upset 5.6% of the sample with an upset then mean score of 2.25 (SD = 1.39) on a 5-point scale, and upset now mean score of 1- not upset (SD = 1.15). Participants were categorized into groups by combining all the child sexual abuse variables together. Group 1 included 11 men (6.8%) who identified themselves as experiencing childhood sexual abuse. Group 2 included 16 men (9.9%) who reported sex with a family member (four of these were also in group 1). Group 3 included 13 men (8.1%) whose sexual acts with non-family members included force, coercion, authority, or a 5-year age difference. Group 4 included seven men (4.3%) who indicated sexually abusing someone 5 or more years younger before they were 18 years old. Group 5 included 128 men who reported no form of sexually abusive acts. While 14.9% of the sample reported only one form of abusive sexual contact, 3.7% experienced two forms of abuse, .6% experienced 3 forms, and 1.2% experienced 4 forms of sexual abuse. An aggregated variable containing both self-identified and researcher-identified abuse yielded 19.3% of the sample experiencing sexually abusive acts. Finally, a severity score 60 was created by summing indicators of abuse (self-identification, family sex, abusive non- family sex, force, frequency, and degree of upset) with 80% of the participants having a CSA severity score of 3 or less. The next highest 10% had severity scores ranging from 4 to 11, and in the highest 10% CSA severity ranged from 12 to 74 points. Childhood Sexual Experiences and Related Variables Several variables from the Childhood Sexual Experiences Checklist (CSEC) were analyzed using bivariate correlations with ? = .05 (2-tailed) as the criterion for significance. The correlations were exploratory in nature to help determine the most potent variables to enter into more complex models. Tables 4 and 5 (next 2 pages) display significant demographic and sexual variables. 61 Table 4. Correlated Demographic and Sexual Variables within the CSEC Variable Correlated Variable Significance p n Father Involvement Father Violence -.311** .000 160 Suicidal Ideation -.168* .034 159 Mental Health Visits -.342** .000 159 CSA Severity -.176* .026 160 A Self-Identification -.178* .024 160 CUSI Forced Adult Sex -.159* .045 160 Father Violence Mother Violence .387** .000 161 Family Sex .233** .004 154 Non-Fam Sex 5 yr Younger .688** .007 14 Force Someone to Have Sex .520** .005 27 CSA Composite .243** .002 161 E with Family .209** .008 161 CSA Perpetration .195* .013 161 CSE Multiple Forms .252** .001 161 Mother Violence CSA Composite .205** .009 161 Suicidal Ideation Mental Health Visits .243** .002 159 Sexual Dissonance .217** .006 160 Unwanted Arousal .193* .014 160 7-day Masturbation Rate .181* .023 151 30-day Masturbation Rate .185* .022 154 CUSI Forced Adult Sex .188* .018 160 Mental Health Visits Sexual Dissonance .325** .000 160 Family Sex .294** .000 153 Pressure Someone for Sex .267** .001 160 Force Someone to Have Sex .472* .013 27 CSA Composite .217** .006 160 A Severity .255** .001 160 CSA Self-Identification .157* .048 160 E with Family .220** .005 160 CSE Multiple Forms .172* .030 160 CUSI Total Score .168* .034 160 CUSI Forced Adult Sex .297** .000 160 CUSI Sex with Children .241** .002 160 Note. CSA = Childhood Sexual Abuse; CUSI = Coping Using Sex Inventory; CSE = Childhood Sexual Experiences; Non- Fam Sex 5 yr Younger = Non-family sex with someone 5 or more years younger; Sexual Dissonance = Sexual experiences contrary to sexual orientation. *p<.05 (2-tailed). **p<.01 (2-tailed). 62 Table 5. Sexualized Coping, Sexual Experiences, and Correlates within the CSEC Variable Correlated Variable Significance p n Sexual Dissonance Suicidal Ideation .217** .006 160 Suicide Attempts .226** .004 160 Mental Health Visits .325** .000 160 Family Sex .237** .003 154 Family Sexual Penetration .515** .000 47 Non-family Sex .158* .045 161 CSA Composite .185* .019 161 A Severity .197* .012 161 CSE with Family .241** .002 161 CUSI Total Score .204** .009 161 CUSI Forced Adult Sex .194* .013 161 Unwanted Arousal Masturbation Under Stress .258** .010 100 Non-fam Sex 5 yr Older .218* .030 99 Non-fam Sex Upset Now .287* .037 53 Pressure Someone for Sex .155* .050 161 CSA Composite .156* .047 161 CSE with Non-family .162* .040 161 CSE Multiple Forms .187* .018 161 Comp. Masturbation 7-day Masturbation Rate .391** .000 159 30-day Masturbation Rate .472** .000 155 Non-family Sex .162* .041 161 CUSI Total Score .354** .000 161 CUSI Sex with Adults .411** .000 161 Mast. Under Stress Non-fam Sex 5 yr Older .422** .000 69 CSA Composite .497** .000 100 A Self-identified .391** .000 100 CSE with Family .211* .035 100 CSE with Non-family .368** .000 100 CSA Perpetration .243* .015 100 CSE Multiple Forms .439** .000 100 7-day Masturbation 30-day Masturbation Rate .924** .000 155 Non-family Forced Sex .247** .007 97 Non-fam sex 5 yr Older .213* .034 99 Pressure for Sex .156* .049 159 CUSI Total Score .270** .001 159 CUSI Sex with Adults .300** .000 159 CUSI Forced Adult Sex .168* .035 159 30-day Masturbation Non-fam Sex 5 yr Older .210* .040 96 CUSI Total Score .317** .000 155 CUSI Sex with Adults .366** .000 155 CUSI Forced Adult Sex .162* .044 155 Note. CSA = Childhood Sexual Abuse; CUSI = Coping Using Sex Inventory; CSE = Childhood Sexual Experiences; Non- Fam Sex 5 yr Younger = Non-family sex with someone 5 or more years younger; Sexual Dissonance = Sexual experiences contrary to sexual orientation; Comp. Masturbation = compulsive masturbation. *p<.05 (2-tailed). **p<.01 (2-tailed). 63 The sample data indicated several forms of sexual experience that could be defined as abusive. To better identify and study these forms of contact, data were coded across several dimensions. A composite variable (CSA Composite) included all forms of sexual experience that could be classified as abusive. Additional distinctions were coded for CSA severity, CSA self-identification, childhood sexual experiences (CSE) with family members and non-family members, CSA perpetration, multiple forms of abuse (CSE Multi), and participants with no CSA experiences. Significant correlations suggest the most relevant components within the groups (Table 6). Table 6. Correlated Sexual Abuse Variables Source (n) All CSA Severity CSA Self ID Family CSE NF CSE CSA Perp Multi CSE Family Sex 154 .678** .767** .319** .930** .203* .232** .676** Family Penetration 47 .339* .438** .389** .445** .406** Family Sex, 5 yr. Older 41 .424** .503** .689** .402** .319* .547** .655** Non-family Sex 161 .164* .171* .160* Non-family Penetration 97 .343** .370** .253* .372** Non-family Forced Sex 97 .246* .245* .220* .210* Non-family Sex, 5 yr. Older 99 .577** .222* .270** .903** .477** .711** Non-family Sex, 5 yr. Younger 14 .576* .571* .782** .826** .832** Upset Now About NF Sex 53 .443** .431** .307* .359** .430** .453** Pressure For Sex 161 .200* .182* Forced Sex 27 .439* Note. Non-significant values omitted. CSA = Childhood Sexual Abuse; CSE = Childhood Sexual Experiences; NF = Non- family; Multi CSE = Multiple forms of childhood sexual experiences r = Pearson correlation. *p<.05 (2-tailed). **p<.01 (2-tailed). 64 Sexualized coping measures. Childhood sexual experiences were significantly related to measures of sexualized coping. Sexualized coping was measured by the Coping Using Sex Inventory (CUSI) in addition to several questions about compulsive masturbation and coercive sexual behavior. The CUSI asked participants to rate 16 statements related to sexual behavior when under stress. A 5-point scale was used to indicate 1- not at all, 2- infrequently, 3- sometimes, 4- often, and 5- almost always the behavior is used whenever the participant encounters a difficult, stressful, or upsetting situation. The CUSI provides a total score, and also 3 factor scores indicating fantasies and behavior related to sex with adults (SWA), forced adult sex (FAS), and sex with children (SWC). The minimum total score is 16 while minimum factor scores are 5 for SWA, 6 for FAS, and 4 for SWC (all items scored 1- not at all). Cronbach?s alpha for the CUSI was ? = .76. The data followed a normal curve for CUSI total score and SWA, but responses for FAS as SWC were highly skewed as most participants denied using these coping strategies while under stress. Table 7 includes summary information for the CUSI. Table 7. Summary Information for Coping Using Sex Inventory (CUSI) Factors Statistic CUSI TOTAL SWA FAS SWC N Valid 160 160 160 160 Missing 1 1 1 1 Mean 28.66 15.98 6.96 4.23 Median 28.50 17.00 6.00 4.00 Std. Deviation 6.34 4.86 1.79 .99 Variance 40.19 23.65 3.23 .97 Percentiles 25 24.00 13.00 6.00 4.00 50 28.50 17.00 6.00 4.00 75 32.00 19.75 7.00 4.00 Note. CUSI = Coping Using Sex Inventory; SWA = Sex with Adults; FAS = Forced Adult Sex; SWC = Sex with Children 65 Bivariate correlations between measures of sexualized coping and CUSI scores indicate behaviors described in the CUSI are significantly correlated with compulsive masturbation, 7-day and 30-day masturbation rates, and pressuring someone for sex. Factor scores for SWA were also significantly related to the same sexual variables. Fantasies and masturbatory behavior related to FAS and SWC were the only scores correlated with participants actually forcing someone to have sex (Table 8). Table 8. Correlations between Sexual Coping Measures Source 2 3 4 5 6 7 8 9 1.Compulsive .391** a .472** b .400** c .442** c Masturbation 2. 7-day ? .924** b .156* a .337** d .347** d .218** d Masturbation 3. 30-day ? .387** e .416** e .211** e Masturbation 4.Pressure ? .301** c .281** c .223** c fo Sex 5. Forced ? .583** f .512** f Sex 6. CUSI ? .918** c .630** c .391** c TotalScore 7. SWA ? .324** c 8. FAS ? .463** c 9. SWC ? Note. Non-significant values omitted. CUSI = Coping Using Sex Inventory; SWA = Sex with Adults; FAS = Forced Adult Sex; SWC = Sex with Children. a n = 159, b n = 154, c n = 160, d n = 155, e n = 151, f n = 27 r = Pearson correlation. *p<.05 (2-tailed), **p<.01 (2-tailed). Attachment to Parents and Friends The Experiences in Close Relationships Inventory. The Experiences in Close Relationships Inventory (ECL) provided a total score and two subscale scores indicating avoidant and anxious attachment patterns. High scores on these subscales indicated difficulty with forming close interpersonal relationships on 66 either anxious or avoidant dimensions. High overall scores indicated more difficulty with attachment. The Avoidance Score ranged from 20 to 120 with a mean of 53.45 (SD = 19.61), the Anxiety Score ranged from 18 to 119 with a mean of 65.53 (SD = 21.05), and the ECL Total Score ranged from 52 to 193 with a mean of 118.98 (SD = 28.32) (Table 9). Cronbach?s alpha for the ECL was ? = .89. Plotting the Avoidance and Anxiety scores, placing a reference line at the mean of each scale and marking each data point as positive (+) or negative (o) for compulsive masturbation produced the scatter plot at Figure 4. Compulsive masturbation did not appear to be significantly related to attachment anxiety or avoidance scores. Table 9. Descriptive Statistics for Experiences in Close Relationships Inventory (ECL) N Mean Median Std. Dev. Range Percentiles Valid Missing 25 50 75 AVOID 161 0 53.45 53.00 19.614 100 38.50 53.00 70.00 ANXIETY 161 0 65.53 64.00 21.049 101 52.50 64.00 81.00 ECLTOTAL 161 0 118.98 120.00 28.315 141 97.50 120.0 137.0 Figure 4. ECL and Compulsive Masturbation ECL Anxiety 120100806040200 E C L A v oi danc e 140 120 100 80 60 40 20 0 Comp. Masturbation 1 0 67 The four quadrants can be used to represent attachment styles (Brennan, Clark, and Shaver,1998). Low Anxiety and low Avoidance represent secure attachment; high Avoidance and low anxiety represent dismissive attachment; high Avoidance and high Anxiety represent fearful attachment; and low Avoidance and high Anxiety represent preoccupied attachment. Participants were classified into attachment groups based on the mean scores for Avoidance and Anxiety, with those at the mean or below in ?low? Anxiety or Avoidance and those with scores above the mean in the ?high? Anxiety or Avoidance. These high and low combinations produced ECL Group1 = Secure, ECL Group 2 = Dismissive, ECL Group 3 = Fearful, and ECL Group 4 = Preoccupied. Using binary logistic regression, ECL group was found to be an unsatisfactory predictor of participants? use of compulsive masturbation with a probability of being correctly classified at 59.6% (due entirely to the constant in the equation). Chi-square for this model was 5.05 (3, n = 161, p = .168). The Inventory of Parent and Peer Attachment. The Inventory of Parent and Peer Attachment asked participant to rate how accurate are statements about mother, father, and peer relationships on a 5-point scale from 1- Almost never true to 5- Almost always true. The scores provide a total attachment score and three subscale scores for Communication, Trust, and Alienation for each category (mother, father, and peers). Higher scores for communication and trust, and low scores for alienation, indicate better attachment relationships. For the total score, alienation scores are reversed so that a high total score reflects more secure attachment while a lower score represents some attachment difficulty. Mean and median scores, standard deviation, and percentiles are provided in Table 10. 68 Table 10. Descriptive Statistics for the Inventory of Parent and Peer Attachment (IPPA) N Mean Median Std. Dev. Range Percentiles Valid Missing 25 50 75 MOTHER 161 0 107.14 109.00 20.847 103 94.50 109.0 125.0 FATHER 161 0 94.06 98.00 25.853 138 82.00 98.00 112.0 PARENTS 161 0 201.19 202.00 40.427 193 178.00 202.0 228.5 PEERS 161 0 85.07 86.00 16.108 98 75.00 86.00 97.00 M_COMM 161 0 37.12 38.00 8.981 40 31.00 38.00 44.00 M_TRUST 161 0 41.90 45.00 8.471 40 40.00 45.00 48.00 M_ALIENA 161 0 17.32 17.00 6.231 30 12.00 17.00 21.00 F_COMM 157 4 31.73 31.00 9.201 40 26.00 31.00 38.00 F_TRUST 157 4 37.54 39.00 8.852 36 33.00 39.00 44.50 F_ALIENA 157 4 23.76 24.00 6.933 28 18.00 24.00 28.00 P_COMM 161 0 28.64 30.00 7.068 40 25.50 30.00 33.00 P_TRUST 161 0 40.50 43.00 8.828 50 38.00 43.00 47.00 P_ALIENA 161 0 16.33 16.00 5.124 33 12.50 16.00 20.00 Note. M_COMM = Mother Communication; M_TRUST = Mother Trust; M_ALIENA = Mother Alienation; F = Father; P = Peer. The means for each IPPA scale and subscale were used as cut points to classify participants into groups with scores below or equal to the mean included in group 1 (Insecure) and those above the mean in group 2 (Secure). Alienation groups were reversed so that low scores were in the secure group and high scores were in the insecure group. These groups were used in a binary logistic regression to determine if IPPA classification could predict compulsive masturbation. None of the subscale scores provided any significant contribution to predicting compulsive masturbation. Next, all IPPA categorical subscales were entered into a regression equation with the 30-day masturbation rate (MAST30) as the dependent variable. Using the backward conditional 69 method of entry (criterion: probability of F-to-remove > = .100), categorical variables for mother communication and trust were modest predictors for 30-day masturbation rates (F 2, 148 = 4.314, p = .015) with an adjusted R 2 = .042. However, when using a binary logistic regression equation to predict attachment categories, MAST30 was only useful in predicting mother communication. The percentage of participants correctly classified using the 30-day masturbation rate was 59.4%, ? 2 (1, n = 155) = 5.525, p = .019. The model accounted for small proportion of variability for mother communication with a Nagelkerke R 2 = .047. The Hosmer-Lemeshow goodness-of-fit test indicated a non- significant difference between predicted and observed classifications with ? 2 (8, n = 155) = 10.008, p = .264. In a similar fashion, the 7-day masturbation rate was used in binary logistic regression equations to predict attachment classifications and was only significant in classifying participants? overall score for parent security. The percentage of correct classifications in this instance was 59.1%, ? 2 (1, n = 159) = 5.059, p = .024. The model accounted for small proportion of variability for parent security with a Nagelkerke R 2 = .042. The Hosmer-Lemeshow goodness-of-fit test indicated a non-significant difference between predicted and observed classifications with ? 2 (6, n = 159) = 11.518, p = .074. Figures 5 and 6 show the mean masturbatory rates for participants according to their degree of mother communication and parental security (1 ? low or insecure, 2 ? high or secure). 70 Figure 5. 30-day Masturbation and Mother Communication MCSECURE 21 M ean M A S T 30 15 14 13 12 11 10 9 Figure 6. 7-day Masturbation and Parent Security PARSEC 21 Me an M A S T 7 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 Plotting the 30-day masturbation rate against the Mother Communication scores produced Figure 7. The reported prior use of compulsive masturbation is indicated by (+) yes and (o) no. Fit lines for the total and subgroups are indicated, showing that those who reported compulsive masturbation showed higher 30-day rates of masturbation and correspondingly lower scores on mother communication. For those men who reported compulsive masturbation, the proportion of variability in their 30-day masturbation rate due to maternal communication was 6%, in contrast to 1% for those who denied compulsive masturbation. In a similar manner, the 7-day masturbation rate was plotted 71 against the overall IPPA Parent attachment score, and compulsive masturbation was indicated in the same manner (Figure 8). Again, those who reported compulsive masturbation indicated a higher 7-day masturbation rate and reported lower overall attachment to their parents. For men who admitted compulsive masturbation, the proportion of variability in 7-day masturbation rates due to parental attachment was 3%, compared to 0% for those who denied compulsive masturbation. Figure 7. Compulsive versus 30-day Masturbation and Mother Communication M_COMM 6050403020100 MA S T 30 70 60 50 40 30 20 10 0 Comp. Masturbation 1 Rsq = 0.0567 0 Rsq = 0.0103 Total Population Rsq = 0.0125 Figure 8. Compulsive versus 7-Day Masturbation and Parent Attachment PARENTS 3002001000 MA S T 7 30 20 10 0 Comp. Masturbation 1 Rsq = 0.0306 0 Rsq = 0.0020 Total Population Rsq = 0.0090 72 Coping Coping in this sample was measured with the Brief-COPE Inventory (B-COPE) and the Coping Using Sex Inventory (CUSI).The B-COPE asked participants to rate twenty-eight statements reflecting the degree to which they applied each coping statement to themselves in the past month. The scale ranged from 1- not at all to 5- very much so. The statements were provided a second time, and participants rated to what degree they typically apply each statement. In this manner a situational coping score and a dispositional coping score was obtained for each participant. Within each score, fourteen subscales were available. Research with the full COPE measure suggested that the 14 scales could be consolidated into three factors, namely Problem-focused Coping, Emotion-focused Coping, and Disengaged Coping (Hudek-Knezevic, Kardum, and Vukmirovic, 1999). However, when the current data were subjected to factor analysis, these factors could not be reproduced from the Brief COPE. While the Brief COPE is a reduction from 4 items per scale to 2 items per scale, there were items added that were not in the full measure, and items in the full measure that were removed from the B-COPE. These small changes may have prevented the formation of problem-focused, emotion-focused, and disengagement factors within this data sample. Cronbach?s alpha for all items of the B-COPE (56 items) was ? = .89, for the 28 situational items was ? = .81, and for the 28 dispositional items was ? = .81. Reliability coefficients for the subscales were smaller with the 14 situational subscales producing ? = .68 and 14 dispositional subscales producing ? = .67 (Cronbach?s alpha). 73 Because of the range of ages in this sample and the potential for age effects on coping behavior, the sample was classified into 3 age groups using K-means clustering procedures. Cluster 1 had one member age 67; Cluster 2 had 151 members with a mean age of 20.51 (SD = 1.91); and Cluster 3 had 9 members with a mean age of 43.56 (SD = 5.41). Limiting the analysis to Cluster 2 members (n = 151), principal components extraction of factors yielded 5 factors with eigenvalues >1. Total variance explained by these factors was 63.21%. The rotated component matrix is shown at Table 11. Reliability was marginal, however, with Cronbach?s alpha for the situational factors at ? = .49. Table 11. Rotated Component Matrix for Situational Coping Situational Component Variable 1 2 3 4 5 S_DENIAL .674 S_SUBST .672 S_VENT .617 S_SFBLAM .577 S_BEHDIS .572 -.452 S_PLANN .853 S_ACTIVE .794 S_INSSUP .803 S_EMSUPP .770 S_RELIG .834 S_SLFDIS .581 S_POSREF .427 .552 S_ACCEPT .784 S_HUMOR .444 .610 Note. Values<.40 are omitted. Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. a. Rotation converged in 10 iterations. b. Only cases for which Cluster Number of Case = 2 are used in the analysis phase. Observing the subscales within each factor, the factors were named according to suggested features contained within them. The situational 5-factor solution included factor 1-Distancing/Avoiding, factor 2- Problem-solving, factor 3- Support-seeking, factor 4- Cognitive Reframing, and factor 5- Acceptance. These factors compare 74 favorably to a 5-factor solution for the COPE suggested by Finch et al. (1999) which included problem-solving, support-seeking, avoidance, reframing, and distancing. Principal components were extracted in a similar manner from the 14 dispositional subscales. Limiting the analysis to members of Cluster 2 (n = 151), only 4 factors emerged with eigenvalues >1 which accounted for only 58.73% of total variance. By requiring a 5-factor solution (factor 5 eigenvalue = .999), the total variance accounted for by those factors was 65.87%. The rotated component matrix is provided in Table 12. Cronbach?s alpha for the dispositional factors was slightly better with ? = .56. Table 12. Rotated Component Matrix for Dispositional Coping Component 1 2 3 4 5 D_ACTIVE .778 D_PLANN .767 D_POSREF .683 D_RELIG .581 D_BEHDIS .802 D_DENIAL .686 -.415 D_SUBST .644 D_EMSUPP .856 D_INSSUP .799 D_SLFDIS .741 D_SFBLAM .726 D_VENT .576 D_ACCEPT .750 D_HUMOR .673 Note. Values <.40 are omitted Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. a. Rotation converged in 8 iterations. b. Only cases for which Cluster Number of Case = 2 are used in the analysis phase. These factors also were named according to subscales contained within them. The five-factor solution for dispositional coping styles included factor 1- Problem-solving, factor 2- Distancing/Avoiding, factor 3- Support-seeking, factor 4- Self-blame/Venting, and factor 5- Acceptance. The structure of these factors was somewhat different from the 75 situational factors, indicating that there may be important differences between participants? situational and dispositional approaches to dealing with problems. Once these 5-factor models were available, correlations were examined between sexual and non-sexual coping. First, sexual behaviors were correlated with situational and dispositional total scores. Then the sexual behaviors were correlated with the 5-factor models, both situational and dispositional forms. Compulsive masturbation and masturbation under stress were not significantly correlated with any coping scale or factor. Both situational and dispositional problem-focused coping bore no significant relationship with any of the sexualized coping behaviors. Dispositional factors were significantly correlated with the highest proportion of sexualized coping behaviors and were therefore selected for model development. Results are shown in tables 13 and 14. Non-significant relationships were omitted from the output. Table 13. Sexualized Coping and the Situational 5-Factor Coping Model Variables Situational Total AVOID PROBLEM SUPPORT REFRAME ACCEPT MAST7 .170* a MAST30 .205* b PRESSURE .224** c .176* c INTER_MA .192* d .193* d .186* d CUSI_TOT .236** e .205** e SWA .211** e FAS .171* e .255** e Note. Non-significant values omitted. MAST7 = 7-day masturbation rate; MAST30 = 30-day masturbation rate; PRESSURE = pressure to gain sex; INTER_MA = number of male sexual partners; CUSI_TOT = Coping Using Sex Inventory total score; FAS = CUSI Forced Adult Sex factor, SWA = CUSI Sex With Adults factor. r = Pearson correlation. a n = 159 b n = 155; c n = 161; d n = 158; e n = 160 *p<.05 (2-tailed), **p<.01 (2-tailed). 76 Table 14. Sexualized Coping and the Dispositional 5-Factor Coping Model Variables Dispositional Total PROBLEM AVOID SUPPORT SELFBLAME ACCEPT MAST7 .165* b MAST30 .182* c PRESSURE .217** a INTER_MA .170* d CUSI_TOT .239** a .222** a .188* a SWA .241** a .170* a .159* a .174* a FAS .221** a .173* a Note. Non-significant values omitted. MAST7 = 7-day masturbation rate; MAST30 = 30-day masturbation rate; PRESSURE = Pressure to gain sex; CUSI_TOT = Coping Using Sex Inventory total score; SWA = CUSI Sex With Adults factor; FAS = CUSI Forced Adult Sex factor. a n = 160; b n = 159; c n = 155; d n = 158 *p<.05 (2-tailed), **p<.01 (2-tailed). Exploring Hypothesis 1 Models of attachment and coping. Understanding the relationships between attachment styles and various coping styles is fundamental to the progress of this research. All attachment subscales (ECL Anxiety and Avoidance, IPPA Communication, Trust, and Alienation for mother, father, and peers) were correlated with dispositional coping total scores and the 5-factor model of coping. All attachment variables were significantly related to at least one dispositional coping variable. Dispositional acceptance was only related to peer attachment variables. Dispositional self-blame/venting was related to anxiety, alienation, and peer trust. Dispositional support-seeking was positively related to communication and trust and negatively related to parents and peer alienation. ECL avoidance was significantly negatively correlated with dispositional problem-solving and support-seeking, and positively correlated with dispositional avoidance. Problem-solving was significantly 77 related to mother and peer communication and trust as well as father communication. It was also negatively correlated with ECL avoidance and mother alienation (Table 15). Table 15. Attachment and the Dispositional 5-Factor Coping Model Dispositional Coping Variables PROBLEM AVOID SUPPORT SELFBLAME ACCEPT ECL -.160* a .170* a -.278** a Avoidance ECL .198* .275** a Anxiety Mother .226** a .265** a Communication Mother .216** a -.179* a .167* a Trust Mother -.171* a .232** a -.199* a .242** a Alienation Father .203* b -.175* b .291** b Communication Father -.224** b .174* b Trust Father .165* b .171* b Alienation Peer .222** a .401** a .234** a Communication Peer .164* a .258** a -.171* a .220** a Trust Peer .209** a -.248** .222** a -.187* a Alienation Note. Non-significant values omitted. r = Pearson correlation. a. n = 161, b. n = 157 *p<.05 (2-tailed), **p<.01 (2-tailed). Hypothesis 1 The first hypothesis states that attachment patterns will predict specific coping strategies in men who have experienced childhood sexual abuse. Preliminary work has served to identify the most potent sexual variables and attachment variables to begin modeling. Additionally, the coping measures were analyzed to help provide detailed 78 understanding of the relationships between sexual and non-sexual coping when under stress. Model of attachment and sexualized coping. Part of Hypothesis 1 rests on the proposed relationship between attachment and sexualized coping behaviors as found in previous research (Lyle, 2003). It was expected that the use of both sexualized and non-sexualized coping would be partly determined by the degree and quality of attachment to mother, father, and friends. Previous researchers (Marshall & Marshall, 2000; Smallbone &McCabe, 2003; Smallbone & Dadds, 1998, 2000) have proposed specific links between parent attachment and sexualized behaviors, and some have suggested links between father attachment and sexual-offending (Marshall & Marshall, 2000; Smallbone & Dadds, 1998, 2000). These sexual behaviors may be shown to correlate with the presence of certain non-sexual coping as well. To test Hypothesis 1, all key variables were entered into a statistical model to test for fit. A Structural Equation Modeling program (AMOS 4.01, SmallWaters Corp., 1994- 1999) was used to build and explore the model, and fit was determined by how closely the model reproduced the observed variables? correlation matrix. The first model included all attachment variables and sexualized coping variables found to be significant in previous correlations. The attachment variables were included as contributors to an unobserved exogenous factor termed ?Attach Problems? and the sexualized coping variables were linked to an unobserved endogenous variable termed ?Sex Coping.? Regression weights for ECL anxiety and CUSI total were held constant at 1.00. The AMOS 4 Regression Weights with corresponding critical values (? = .05) and the Modification Indices were used to systematically remove variables and link error terms. 79 Fit was determined by the following guidelines: ? 2 /df<2; GFI>.90; and RMSEA<.05 with p>.05. Several variations of the model were explored with few reaching the threshold for significance. Most attachment variables remained in the model except ECL Avoidance, Father Alienation, Peer Communication, and Peer Alienation. The greatest difficulty in producing a significant model at this stage was selecting the most powerful sex coping variables. Modification indices suggested the best sex coping variables were the CUSI Forced Adult Sex (FAS) with regression weight held constant at 1.00, MAST30, and Pressure as indicators of sexualized coping. Because age effects had been detected in other analyses, the model was restricted to those men in Cluster 2 (those ranging from 19 to 29 years old, mean age 21, n = 151). All models required linking several of the error terms to improve the models? fit with the observed correlation matrix. The first model of attachment and sexualized coping is at Figure 9. For Model 1, ? 2 = 50.912 with 28 degrees of freedom, yielding ? 2 /df = 1.818, p = .005; the Goodness of Fit (GFI) = .936; the Normed Fit Index (NFI) = .92; and the Root Mean Squares Error of Approximation (RMSEA) = .074, p = .111. The direct effect of attachment on sex coping in this model (using standardized path coefficients) was b = .39 (p<.05) with R 2 = .15. Table 16 provides standardized regressions weights and p-values for Model 1. 80 Figure 9. Model 1 ? Attachment and Sexualized Coping, Age-restricted (n = 151) Attach Problems .15 Sex Coping .23 ECL Anxiety .80 Mother Trust .10 Father Comm .10 Peer Trust .23 Father Trust .82 Mother Alienation .67 Mother Comm .14 CUSI FAS e1 e3 e4 e5 e7 e6 e10 e12 .39 -.48 .37 e19 -.32 .77 .41 MAST30 .11 PRESSURE e13 e15 .64 .33 -.25 .33 .18 Chi-Sq = 50.912 df = 28 Chi-Sq/df = 1.818 GFI = .936 Adjusted GFI = .874 NFI = .918 RMSEA = .074 p = .111 -.89 .90 -.82 .16 -.13 -.31 .47 Interpreting Model 1, attachment problems (high anxiety, low parental/peer communication and trust, high alienation) are positively correlated with sexualized coping behaviors including masturbation and fantasies/behavior directed toward forced adult sex, high rates of masturbation per month, and pressuring someone to have sex. 81 Table 16. Model 1 Regression Weights Model 1 Std. Coeff. S.E. C.R. P Sex Coping <-- Attach Problems 0.392 0.015 2.274 0.023 ANXIET <-- Attach Problems 0.475 M_COMM <-- Attach Problems -0.818 0.159 -4.932 0.000 M_TRUST <-- Attach Problems -0.892 0.161 -5.158 0.000 M_ALIENA <-- Attach Problems 0.905 0.111 5.323 0.000 F_COMM <-- Attach Problems -0.320 0.091 -3.174 0.002 F_TRUST <-- Attach Problems -0.483 0.101 -4.032 0.000 P_TRUST <-- Attach Problems -0.311 0.088 -3.219 0.001 FAS <-- Sex Coping 0.371 MAST30 <-- Sex Coping 0.641 2.792 2.343 0.019 PRESSURE <-- Sex Coping 0.326 0.063 2.257 0.024 Note. Model 1 n = 151 Attach = Attachment; Sex Coping = Sexualized Coping; ANXIET = ECL Anxiety; M_TRUST = Mother Trust; M_COMM = Mother Communication; M_Aliena = Mother Alienation; F_TRUST = Father Trust; F_COMM = Father Communication; P_TRUST = Peer Trust; MAST30 = 30-day Masturbation Rate; FAS = CUSI Forced Adult Sex; PRESSURE = Pressuring someone for sex. Modeling attachment and coping. After establishing the relationship between attachment and sexualized coping, the next step was to create a model of attachment and coping. All attachment variables were linked to an exogenous variable termed ?Attach Problem? and all coping variables were linked to an endogenous variable termed ?Coping.? The analysis was restricted to Cluster 2 , (n = 151) because of anticipated age effects on coping behavior. The initial model indicated that the three sex coping variables were not related to the coping variables. Additionally, dispositional avoidance (D5Av) was not related to the coping variables for this sample (Figure 10, Model 2). 82 Figure 10. Model 2 ? Exploring Attachment and Coping Attach Problems .11 AV e1 .33 .19 AX e2 .43 .66 MC e3 -.81 .71 MT e4 -.85 .87 MA e5 .93 .13 FC e6 -.36 .25 FT e7 -.50 .15 FA e8 .38 .03 PC e9 -.18 .11 PT e10 -.33 .11 PA e11 .33 .50 D5P e19 .01 D5Av e18 .61 D5S e17 .06 D5SB e16 .19 D5Ac e15 .00 M30 e14 .00 Pr e13 .00 FAS e12 Chi-Sq = 598.055 df = 148 Chi-Sq/df = 4.041 p = .000 GFI = .693 Adjusted GFI = .606 NFI = .534 RMSEA = .142 p = .000 .08 .25 .76 e20 .14 Coping -.38 -.05 .00 -.03 .43 .24 .78 -.11.71 Indicators of fit suggest Model 2 does not adequately reproduce the observed correlation matrix, but as an exploratory model, establishes that the sex coping variables (FAS, Pr, and M30) and dispositional avoidance (D5Av) are not related to the other coping variables. Thus, the coping variables were divided into two (see Figure 11), placing Problem-focused (D5P), Support-seeking (D5S), Self-Blame/Venting (D5SB), and Acceptance D5Ac) on one side termed ?Non-sex Coping.? On the other side, sex coping variables included those already identified as important in this sample including 30-day masturbation rate (M30), Pressuring someone for sex (Pr), and CUSI Forced AV = ECL Avoidance AX = ECL Anxiety MC = Mother Communication MT = Mother Trust MA = Mother Alienation FC = Father Communication FT = Father Trust FA = Father Alienation PC = Peer Communication PT = Peer Trust PA = Peer Alienation FAS = CUSI Forced Adult Sex Pr = Pressure M30 = Masturbation 30-days D5Ac = Dispositional Acceptance D5SB = Dispositional Self-blame/Venting D5S = Dispositional Support-seeking D5Av = Dispositional Avoidance D5P = Dispositional Problem-focused 83 Adult Sex (FAS). In addition, the dispositional coping variable, Avoidance (D5Av) was included on this side as it did not relate to the non-sexual coping variables. Regression weights for attachment anxiety (AX), Problem-focused coping (D5P), and 30-day masturbation rate (M30) were held constant at 1.00. The AMOS 4 Modification Indices suggested several correlations between attachment scale error terms. The standardized correlation coefficient between attachment and sex coping was b = .38 (p<.05), R 2 = .14 and between attachment and non-sex coping was b = -.35 (p<.01), R 2 = .12 (see Figure 11). Next, the CSA Composite variable (CSA_COMP) was entered into the model. This is the variable that aggregated all forms of child sexual experiences that could be considered abusive. The CSA variable was significantly correlated with attachment for this sample (b = .19, p<.05), but accounted for only a small portion of the variability in attachment (R 2 = .04). Model 3 (Figure 11) demonstrates these relationships, and while the fit estimates are significant, the p-value for RMSEA is not yet >.05. 84 Figure 11. Model 3 ? CSA, Attachment, and Coping .04 Attach Problems .12 Non-Sex Coping .10 AV e1 .32 .18 AX e2 .42 .67 MC e3 -.82 .75 MT e4 -.87 .84 MA e5 .92 .15 FC e6 -.38 .30 FT e7 -.55 .17 FA e8 .42 .02 PC e9 .11 PT e10 -.33 .09 PA e11 .29 .51 D5P e19 .72 .22 D5Ave18 .58 D5S e17 .76 .07 D5SB e16 .20 D5Ac e15 .45 .23 M30e14 .14 Pre13 .21 FASe12 Chi-Sq = 376.097 df = 158 Chi-Sq/df = 2.380 p = .000 GFI = .804 Adjusted GFI = .739 NFI = .713 RMSEA = .096 p = .000 .06 .69 e20 .15 Sex Coping .47 .45 .38 .48 e21 CSA_COMP e22 .19 .68 -.32 .29 -.06 .17 -.29 .35 -.16 .26 -.15 .38 -.35 Model 3 supports the first hypothesis by indicating there is a significant relationship between abusive childhood sexual experiences and the quality of attachment to mother, father, and friends. In turn, attachment plays a significant role in determining the type of coping behaviors typically used when under stress. Poor attachment predicts increased sexualized coping and avoidance strategies, and also predicts decreased non- sexualized coping. Model 3 regression weights are shown in Table 17. 85 Table 17. Model 3 Regression Weights Model 3 Std. Coeff. S.E. C.R. p Attach Problem <-- CSA_COMP 0.193 2.013 2.131 0.033 Non-Sex Coping <-- Attach Problem -0.352 0.072 -2.916 0.004 Sex Coping <-- Attach Problem 0.381 0.098 2.436 0.015 AVOID <-- Attach Problem 0.319 0.212 3.196 0.001 ANXIET <-- Attach Problem 0.420 M_COMM <-- Attach Problem -0.820 0.158 -5.064 0.000 M_TRUST <-- Attach Problem -0.865 0.160 -5.211 0.000 M_ALIENA <-- Attach Problem 0.918 0.121 5.280 0.000 F_COMM <-- Attach Problem -0.384 0.113 -3.628 0.000 F_TRUST <-- Attach Problem -0.545 0.115 -4.414 0.000 F_ALIENA <-- Attach Problem 0.415 0.089 3.779 0.000 P_COMM <-- Attach Problem -0.145 0.067 -1.647 0.100 P_TRUST <-- Attach Problem -0.333 0.080 -3.589 0.000 P_ALIENA <-- Attach Problem 0.293 0.048 3.499 0.000 D5PROBLM <-- Non-Sex Coping 0.716 D5SPPORT <-- Non-Sex Coping 0.764 0.124 5.306 0.000 D5SLFBLM <-- Non-Sex Coping 0.260 0.098 2.679 0.007 D5ACCEPT <-- Non-Sex Coping 0.448 0.074 4.404 0.000 MAST30 <-- SexCoping 0.478 FAS <-- SexCoping 0.454 0.052 2.800 0.005 PRESSURE <-- SexCoping 0.379 0.008 2.590 0.010 D5AVOID <-- SexCoping 0.471 0.156 2.830 0.005 Note. n = 151. Attach = Attachment; Sex Coping = Sexualized Coping; AVOID = ECL Avoidant; ANXIET = ECL Anxiety; M_TRUST = Mother Trust; M_COMM = Mother Communication; M_ALIENA = Mother Alienation; F_TRUST = Father Trust; F_COMM = Father Communication; F_ALIENA = Father Alienation; P_TRUST = Peer Trust; P_COMM = Peer Communication; P_ALIENA = Peer Alienation; D5PROBLM = Problem-focused Coping; D5SPPORT = Support-seeking; D5SLFBLM = Self-blaming/venting; D5ACCEPT = Acceptance; D5AVOID = Avoidance; MAST30 = 30-day Masturbation Rate; FAS = CUSI Forced Adult Sex; PRESSURE = Pressuring someone for sex Several of the observed variables provided smaller contributions to the latent variables and were systematically removed to improve model fit. Removing the coping variable D5Self-Blame/Venting (D5SB) and Peer Attachment variables (PC, PT, & PA) improved the model?s fit to the observed correlation matrix. The revised model is shown at Figure 12, Model 4. 86 Figure 12. Model 4 ? CSA, Attachment, and Coping (Revised Model) .03 Attach Problems .13 Non-Sex Coping .24 AX e2 .49 .69 MC e3 -.83 .80 MT e4 -.89 .81 MA e5 .90 .10 FC e6 -.32 .24 FT e7 -.49 .15 FA e8 .39 .52 D5P e19 .72 .21 D5Ave18 .57 D5S e17 .75 .19 D5Ac e15 .44 .24 M30e14 .14 Pre13 .21 FASe12 Chi-Sq = 119.334 df = 81 Chi-Sq/df = 1.473 p = .004 GFI = .907 Adjusted GFI = .863 NFI = .864 RMSEA = .056 p = .305 .25 .77 e20 .15 Sex Coping .46 .45 .38 .49 e21 CSA_COMP e22 .19 -.57 .24 .38 -.36 .35 -.48 .17 Bold = Significant at p<.05 or better Italic = Non-significant Removing the peer attachment variables and dispositional self-blame from the model provides a much better fit to the observed correlation matrix. Model 4 demonstrates that childhood sexual abuse contributes to attachment problems and is significantly related to anxiety, maternal attachment, and paternal attachment. These attachment problems are predictive of diminished use of non-sexualized coping, namely problem-focused coping, support-seeking, and acceptance; and predictive of increased dispositional avoidance and sexualized coping behavior. Attachment qualities predict specific coping behaviors. Hypothesis 1 also proposed that attachment quality (anxious or avoidant) and attachment focus (mother, father, and peers) would predict certain combinations of coping behaviors. Men with Avoidant attachment styles would rely on masturbation and 87 disengagement strategies when under stress. Men with Anxious attachment styles would predominantly use frequent masturbation and consensual sex when under stress, and engage in more emotion-focused coping strategies when compared to less anxious individuals. Poor maternal attachment would predict fantasies and behavior related to forced adult sex, poor paternal attachment would predict sex directed toward children, and poor peer attachment would predict more frequent use of coercion and pressure for sex. In general, more secure individuals would possess internal working models of themselves as competent, and would therefore utilize more problem-focused strategies, support-seeking, and reframing compared to less secure individuals. Several procedures were involved in identifying the variables that might be most potent in each attachment category. First, the sample was restricted to Cluster 2 (n = 151) with a mean age of 21 years. As before, age effects were indicated in the attachment and coping of the sample, and restricting the age would produce a more reliable estimation of variable relationships. High scores for avoidance and anxiety were indicative of problems with attachment while high scores for mother, father, or peers were indicative of attachment security. In addition, four ECL attachment groups (created by combining high and low avoidance and anxiety scores) were available for further analysis of coping behaviors. These groups were group 1 ? Secure, group 2 ? Dismissive, group 3 ? Fearful, and group 4 ? Preoccupied. Next, linear regressions were created to examine the predictive value of several sex coping variables and dispositional coping variables in each attachment category. The first regression placed Avoidance as the dependent variable. The sexual variables (Comp. Masturbation, Mast. Under Stress, Mast30, Mast7, Pressure, and CUSI SWA, FAS, and 88 SWC) were selected for the first block with the stepwise method of entry (criteria: probability-of-F-to-enter < = .050, probability-of-F-to-remove > = .100). In the second block, 5 dispositional coping variables (D5Problem, D5Avoid, D5Support, D5SlfBlm, and D5Accept) were selected with a stepwise method of entry. Missing data were replaced with the mean. Using only Cluster 2 participants, five regressions were run using each (continuous) attachment variable (Avoidance, Anxiety, Mother, Father, and Peers). Table 18 provides summary information for the regressions. To make specific predictions by ECL attachment group, the set of regressions were run again with mother, father, and peer attachment as the dependent variable, and ECL groups 1-4 as the selection variable. Summary information appears in Table 19. Table 18. Sexual and Non-sexual Predictors of Attachment (n = 151) Model Source Predictors R 2 F Std. Coeff. t Avoidance D5Support -.280 -3.631*** D5Avoid .13 10.863 .196 2.543* Anxiety D5Self Blame .301 3.766*** D5Support .10 8.141 -.176 -2.211* Mother 30-day Masturbation -.299 -3.727*** SWA .291 3.612*** D5Support .206 2.364* D5Self Blame -.184 -2.434* D5Problem .24 8.887 .186 2.137* Father FAS -.115 -1.509 D5Support .367 4.768*** D5Self Blame .17 10.347 -.219 -2.812** Peers D5Support .08 12.283 .276 3.505** Note. High Avoidance and Anxiety scores indicate problematic attachment; High Mother, Father, and Peer scores indicate more secure attachment (Low parent and peer scores indicate poor attachment). Std. Coeff. = Standardized Coefficient ? Beta; SWA = CUSI Sex with Adults; FAS = CUSI Forced Adult Sex; D5 = Dispositional 5-factor model of Coping; Support = Support-seeking; Problem = Problem-focused coping; Avoid = Avoidance coping * p<.05; ** p<.01; *** p<.001 89 Table 19. Sexual and Non-sexual Predictors of Attachment within ECL Groups (n = 151) Model Source ECL Group Predictors R 2 F Std. Coeff. t Mother Secure D5Self Blame .13 6.135 -.361 -2.416* Dismissive 30-day Masturbation .25 10.225 -.512 -3.212** Fearful SWA .13 5.299 .363 2.235* Preoccupied SWA .18 7.940 .420 2.741* Father Secure ? ? ? ? ? Dismissive ? ? ? ? ? Fearful FAS .17 7.210 -.413 -2.607* Preoccupied D5Support .558 3.785** D5Self Blame .32 8.410 -.334 -2.264* Peers Secure ? ? ? ? ? Dismissive Pressure .13 4.427 -.359 -2.069* Fearful Mast. Under Stress .12 4.881 -.350 -2.209* Preoccupied FAS .236 1.638 D5Support .28 7.028 .430 2.978** Note. High Mother, Father, and Peer scores indicate more secure attachment (Low parent and peer scores indicate poor attachment). Beta = Standardized Coefficient ? Beta; Adj. R 2 = Adjusted R 2 ; Pressure = Pressuring someone for sex; SWA = CUSI Sex with Adults; FAS = CUSI Forced Adult Sex; D5 = Dispositional 5-factor model of Coping; Support = Support-seeking; Problem = Problem-focused coping; Avoid = Avoidance coping * p<.05; ** p<.01; *** p<.001 In the Cluster 2 sample, avoidant or anxious attachment were not predictive of any specific sexualized coping. Highly avoidant men used less support-seeking and more avoidance coping. Highly anxious men reported more self blame/venting and less support-seeking. Secure maternal attachment was predicted by lower rates of masturbation under stress, relatively more fantasies and behavior related to consensual sex, less self blame/venting, and more problem-focused coping. Secure Paternal attachment was predicted by fewer tendencies to engage in fantasies and behavior related to forced sex, more support-seeking, and less self-blame/venting. Secure peer attachment was predictive of increased support-seeking. 90 Examining ECL attachment groups, within the Dismissive attachment group, insecure maternal attachment predicted higher masturbation rates and insecure peer attachment predicted higher tendency to pressure someone for sex. In the Fearful group, insecure maternal attachment was predictive of lower scores for CUSI Sex with Adults, insecure paternal attachment predicted higher scores for CUSI Forced adult Sex, and insecure peer attachment predicted increased masturbation under stress. In the Preoccupied group, insecure maternal attachment predicted lower scores on CUSI Sex with Adults. No other sexualized coping was predictive of attachment scores among the preoccupied men. Secure men did not demonstrate significant sexual behavior predictive of any attachment score. Similar linear regressions were repeated, limiting the sample to the CSA Composite group, but there were no sexualized coping predictors of attachment within the abused group. The specific predictions proposed in the first hypothesis relating attachment styles and coping behaviors were not entirely supported, in part due to the unavailability of the 3-factor representation of coping. However, coping researchers (Lazarus & Folkman, 1989; Carver et al. 1989) describe problem-focused behavior as those that imply something can be done, while emotion-focused strategies imply reducing or managing the distress. Tamres et al. (2002) suggest emotion-focused strategies include venting, ruminating, avoidance, acceptance, reinterpretation, and self-blame. In contrast, problem-focused behaviors include active planning, suppression of competing activities, restraint, and support-seeking. Within this framework, the men did not report the expected sexualized coping related to avoidant and anxious attachment. There was a tendency to report less problem-focused and support-seeking behavior and more avoidant 91 behavior. When attachment was classified within the two dimensions (anxious and avoidant) into four attachment groups (Brennan, Clark, and Shaver, 1998) results indicated some sexualized coping was predictive of dismissive and fearful attachment. Insecure maternal attachment predicted some sexualized coping, including increased masturbation and decreased interest in consensual sexual themes, but the expected interest in forced adult sex was not supported. Insecure paternal attachment predicted increased interest in fantasies or behavior related to forced sex, but peer attachment predicted no form of sexualized coping. Hypothesis 2 The second hypothesis proposed that insecure men who recognized they had been abused would use more problematic dispositional coping strategies, including disengagement and sexualized coping, than their securely attached peers. SEM procedures were used to test the relative use of certain coping behaviors within the subset of men who identified their childhood sexual experiences as abusive. Model 5 (Figure 13) below indicates sexualized coping behaviors on the left side and non-sexualized coping behaviors on the right. This division of coping scales was based upon earlier indications that the sex coping variables and dispositional avoidance coping were not related to the other dispositional coping scales (see Figure 12, Model 4). For those men who self- identified (CSA Self-ID), there is a significant relationship between CSA and both categories of coping. Before entering the effect of attachment problems, the direct effect of self-identified abuse (CSA SID) on sexualized coping (using standardized path coefficients) is b = .65 (p<.01) with R 2 = .42. In this model, sexualized coping relates predominantly to Masturbation Under Stress, while the other indicators account for little 92 or no variability in the model. Independent of the effect of attachment problems, the direct effect of CSA SID on non-sexual coping is b = -.22 (p<.05) with R 2 = .05. Avoidance (on the left) is not significantly related to sexualized coping for men who self- identify CSA. From the proportion of variability in sex coping and non-sex coping, it appears that CSA SID is more important in relation to sex coping than non-sex coping, indicating that for those men who recognize their experiences as abusive, there is a stronger tendency to use sexualized and avoidant coping under stress rather than non- sexual coping. There was no significant direct relationship between CSA SID and Attachment. The introduction of Attachment to the model produced negligible changes in the direct effect of self-identified child sexual abuse on sexualized coping. However, attachment problems decreased the use of non-sexual coping to a significant degree (b = -.39) and the combined effect of attachment problems and abuse accounted for 19% of the variability in non-sexual coping scores. Although the variables of interest showed significant relationships in the full model (all attachment and coping variables included), the p-value for the root mean squared error of approximation (RMSEA) did not exceed the cut point (>.05), indicating this model did not completely reproduce the observed correlation matrix. For that reason, the three Peer attachment variables, Dispositional Avoidance (D5Av) and Dispositional Self-Blame/Venting (D5SB) were removed from the model, greatly improving the fit (Figure 13, Model 5) with ? 2 = 123.614, p = .001, ? 2 /df = 1.565, GFI = .904, NFI = .861, RMSEA = .061. p = .181. 93 Figure 13. Model 5 ? Self-identified Abuse, Attachment, and Coping .00 Attach Problems .19 Non-Sex Coping .23 AX e2 .48 .70 MC e3 -.83 .79 MT e4 -.89 .81 MA e5 .90 .11 FC e6 .24 FT e7 .15 FA e8 .49 D5P e19 .70 .58 D5S e17 .76 .21 D5Ac e15 .46 .30 M ST e14 .08 Pre13 .05 FASe12 Chi-Sq = 123.614 df = 79 Chi-Sq/df = 1.565 p = .001 GFI = .904 Adjusted GFI = .854 NFI = .861 RMSEA = .061 p = .181 .24 .77 e20 .49 Sex Coping .23 .29 e21 CSA Self-ID e22 -.57 .24 .55 -.33 -.49 .39 .70 -.23 -.48 .10 -.38 .08 AV e1 .29 -.06 Bold = Significant at p<.05 or better Italic = Non-significant .32 .17 As predicted in Hypothesis 2, Model 5 indicates that men who self-identified experiencing sexual abuse were more likely to use masturbation during periods of stress than other forms of coping. Self-identified CSA increased the likelihood of using sexualized coping and decreased the likelihood of using non-sexualized coping. Attachment problems had a significant effect on the use of non-sexual coping strategies, with secure men using more non-sexual coping and insecure men using fewer non- sexualized coping strategies. Model 5 regression estimates are provided at Table 20. 94 Table 20. Model 5 Regression Estimates Model 5 Std. Coeff. S.E. C.R. p Attach Problems <-- CSA_SID -0.059 3.475 -0.692 0.489 Sex Coping <-- CSA_SID 0.696 0.123 5.061 0.000 Non-Sex Coping <-- CSA_SID -0.234 1.955 -2.501 0.012 Sex Coping <-- Attach Problems 0.101 0.003 0.783 0.433 Non-Sex Coping <-- Attach Problems -0.375 0.061 -3.135 0.002 AVOID <-- Attach Problems 0.286 0.176 3.040 0.002 ANXIET <-- Attach Problems 0.478 M_COMM <-- Attach Problems -0.834 0.133 -5.338 0.000 M_TRUST <-- Attach Problems -0.888 0.138 -5.427 0.000 M_ALIENA <-- Attach Problems 0.899 0.096 5.679 0.000 F_COMM <-- Attach Problems -0.326 0.091 -3.342 0.001 F_TRUST <-- Attach Problems -0.489 0.096 -4.414 0.000 F_ALIENA <-- Attach Problems 0.389 0.073 3.777 0.000 D5PROBLM <-- Non-Sex Coping 0.701 D5SPPORT <-- Non-Sex Coping 0.761 0.128 5.245 0.000 D5ACCEPT <-- Non-Sex Coping 0.455 0.077 4.419 0.000 FAS <-- Sex Coping 0.226 0.939 1.972 0.049 PRESSURE <-- Sex Coping 0.287 0.174 2.400 0.016 MAST_STR <-- Sex Coping 0.548 Note. N = 151. Sex Coping = Sexualized Coping; AVOID = ECL Avoidance; ANXIET = ECL Anxiety; M_TRUST = Mother Trust; M_COMM = Mother Communication; M_ALIENA = Mother Alienation; F_TRUST = Father Trust; F_COMM = Father Communication; F_ALIENA = Father Alienation; P_TRUST = Peer Trust; P_COMM = Peer Communication? P_ALIENA = Peer Alienation; D5PROBLM = Problem-focused Coping; D5SPPORT = Support-seeking; D5SLFBLM = Self-blaming/venting; D5ACCEPT = Acceptance; D5AVOID = Avoidance; MAST_STR = Masturbation under stress; FAS = CUSI Forced Adult Sex; PRESSURE = Pressuring someone for sex Hypothesis 3 The third hypothesis proposed that if attachment was significantly and directly related to CSA and coping strategies, it would also mediate the relationship between CSA and coping. The basic model from Hypothesis 1 was utilized to test the effect of attachment on coping among men who had childhood sexual experiences. The models are presented in series. The first view (Figure 14, Model 6) demonstrates the effect that 95 attachment exerts in sex-coping and non-sex coping for the CSA Composite group (n = 27). There is a significant relationship (b = .19, p<.05) between all forms of sexual abuse (CSA Comp) and Attachment. Attachment is also significantly related to Sex Coping (b = .38, p<.05) and Non-sex Coping (b = -.36, p<.01). While CSA accounts for a very small proportion of variability in Attachment (R 2 = .03), Attachment accounts for a modest proportion of Sex Coping (R 2 = .15) and Non-sex Coping (R 2 = .13). Figure 14. Model 6 ? Relationship of CSA to Attachment and Coping .03 Attach Problems .13 Non-Sex Coping .24 AX e2 .49 .69 MC e3 -.83 .80 MT e4 -.89 .81 MA e5 .90 .10 FC e6 -.32 .24 FT e7 -.49 .15 FA e8 .39 .52 D5P e19 .72 .21 D5Ave18 .57 D5S e17 .75 .19 D5Ac e15 .44 .24 M30e14 .14 Pre13 .21 FASe12 Chi-Sq = 119.334 df = 81 Chi-Sq/df = 1.473 p = .004 GFI = .907 Adjusted GFI = .863 NFI = .864 RMSEA = .056 p = .305 .25 .77 e20 .15 Sex Coping .46 .45 .38 .49 e21 CSA_COMP e22 .19 -.57 .24 .38 -.36 .35 -.48 .17 Bold = Significant at p<.05 or better Italic = Non-significant 96 Model 6 (above) indicates that men who have had abusive childhood sexual experiences and poorer attachment bonds are more likely to use sexual and avoidant coping strategies, while men with more secure bonds are more likely to use non-sexual coping strategies when under stress. Within the sexual coping strategies, the four identified behaviors have approximately equal factor loadings on sex coping. On the non- sex coping side of the model, support-seeking and problem-solving strategies have highest factor loadings on non-sex coping variable, followed by acceptance coping. Self- blame is also significantly related to non-sex coping, but provides a relatively small contribution to non-sexual coping in this configuration; therefore it was removed. It may be interpreted, therefore, that one who has poor attachment bonds is less likely to use problem-focused and support-seeking coping strategies when under stress. We can also examine the factor loadings among the attachment variables and see that mother attachment has the strongest relationship with the coping behaviors of these men. The next figure (Figure 15, Model 7) demonstrates the direct effect of CSA on Sex-coping and Non-sex Coping. This figure shows that the Composite CSA variable is significantly related to Sex Coping (b = .33, p<.05, R 2 = .11) and Non-sex Coping (b = -.27, p<.01, R 2 = .08), and remains significantly correlated to attachment. Thus, the requirements for testing mediation are met with significant relationships in all direct links of the variables. 97 Figure 15. Model 7 ? Direct Effects of CSA on Coping and Attachment .03 Attach Problems .08 Non-Sex Coping .24 AX e2 .69 MC e3 -.83 .80 MT e4 -.89 .80 MA e5 .89 .10 FC e6 .24 FT e7 .15 FA e8 .51 D5P e19 .71 .20 D5Ave18 .55 D5S e17 .74 .21 D5Ac e15 .46 .13 M30e14 .18 Pre13 .30 FASe12 Chi-Sq = 125.620 df = 81 Chi-Sq/df = 1.551 p = .001 GFI = .906 Adjusted GFI = .860 NFI = .856 RMSEA = .061 p = .195 .27 .77 e20 .11 Sex Coping .55 .42 e21 CSA_COMP e22 -.57 .26 .36 -.32 -.49 .39 .45 .18 Bold = Significant at p<.05 or better Italic = Non-significant .33 -.27 .49 .36 -.48 .18 In Model 7 above, there is a significant relationship between abusive childhood sexual experiences and the coping behaviors used when under stress. According to the diagram, the presence of abusive sexual contact in childhood or early adolescence is positively related to sexualized and avoidant coping, and negatively related to more appropriate non-sexual coping. 98 Exploring abuse indicators of attachment and coping. Additional exploration of this model was undertaken, replacing the CSA Composite variable with several other indicators of sexual abuse to examine any direct effects between the indicator and attachment and coping, and provide a more detailed conceptualization of the potent elements of abuse. Few of the indicators had a direct significant relationship with attachment. Several CSA groups and abuse indicators were significantly related to sex coping and/or non-sex coping, as shown in Table 21. Table 21. Abuse Indicators Predicting Attachment Problems, Sex Coping, and Non-sex Coping Standardized Regression Weights Predictors Attachment Problems Sex Coping Non-sex Coping CSA Groups CSA Composite Group (Model 7) .18* .33* -.26** CSA Self-Identified -.22* CSA Family .32* CSA Non-family .32* -.36*** CSA Multi .18* .40* -.37*** CSA None -.19* -.36* .21* Abuse Indicators Upset Now a .39* -.26** Family Sex .37* Family Penetration .56** Family Sex with someone 5 years older -.74*** Non-family Penetration .34* -.26** Non-family Forced Sex .37** Non-family Sex with someone 5 years older .34** -.30** Upset Now b .52* -.29** Note. n = 151 (Cluster 2 members) Variables were entered individually into Model 7 to provide standardized regression weights. a Upset Now about CSA before age 13; b Upset now about Non-family Sex * p<.05; ** p<.01; *** p<.001 99 Mediation of the abuse ? attachment relationship. The third hypothesis predicted attachment would mediate the relationship between abuse and coping. The test of mediation requires the direct effect of CSA on coping and attachment (established by Model 7) and that the introduction of the relationship between attachment and coping reduces the effect of CSA on coping to non- significance or zero. The model for this relationship is below at Figure 16, Model 8. Figure 16. Model 8 ? Mediation: CSA, Attachment, and Coping .03 Attach Problems .17 Non-Sex Coping .24 AX e2 .49 .69 MC e3 -.83 .80 MT e4 -.89 .80 MA e5 .90 .10 FC e6 -.32 .24 FT e7 -.49 .15 FA e8 .39 .50 D5P e19 .71 .23 D5Ave18 .57 D5S e17 .76 .20 D5Ac e15 .45 .18 M30e14 .15 Pre13 .25 FASe12 Chi-Sq = 108.648 df = 79 Chi-Sq/df = 1.375 p = .015 GFI = .916 Adjusted GFI = .872 NFI = .876 RMSEA = .050 p = .480 .26 .77 e20 .20 Sex Coping .48 .50 .39 .42 e21 CSA_COMP e22 .18 -.57 .24 .27 -.22 .31 -.32 Bold = Significant at p<.05 or better Italic = Non-significant .36 -.48 .17 100 Model 8 demonstrates that attachment diminished the direct effect of CSA on both coping variables, but the strength of the direct effect of CSA on Sex Coping remained significant (b = .27, p<.05) and the effect of CSA on Non-sex Coping also remained significant (b = -.22, p<.05). Thus, contrary to expectation, current attachment did not mediate the effect of childhood sexual experiences on coping behaviors in this sample. When all of the other significant indicator variables were substituted in this model, attachment did not mediate any form of coping. One exception occurred when those with no abuse experiences were entered into the model. In that single case, attachment significantly mediated the use of non-sexualized coping, so that men who had not been abused, but who had poor attachment relationships were less likely to use non- sexual (adaptive) means of coping when under stress than those who were more securely attached. The directionality of the attachment relationship was then reversed so that CSA influences Coping, and Coping influences Attachment. The direct effect of CSA on Attachment was reduced to nearly zero. Model 9 (Figure 17) below demonstrates these relationships. 101 Figure 17. Model 9 ? Mediation of the Direct Effect of CSA on Attachment .33 Attach Problems .07 Non-Sex Coping .31 AX e2 .55 .47 MC e3 -.69 .37 MT e4 -.61 .57 MA e5 .75 .22 FC e6 -.47 .37 FT e7 -.60 .20 FA e8 .44 .46 D5P e19 .68 .25 D5Ave18 .62 D5S e17 .79 .20 D5Ac e15 .44 .16 M30e14 .15 Pre13 .24 FASe12 Chi-Sq = 112.471 df = 78 Chi-Sq/df = 1.442 p = .006 GFI = .911 Adjusted GFI = .863 NFI = .872 RMSEA = .054 p = .359 .14 .73 e20 .11 Sex Coping .50 .49 .39 .41 e21 CSA_COMP e22 .00 -.55 .14 .33 -.27 .37 -.40 (NS) p=.02 p=.01 p=.04 p=.00 Bold = Significant at p<.05 or better Italic = Non-significant .55 -.67 -.47 -.42 This mediational model (Model 9 above) shows that CSA has a direct effect on both forms of coping, and coping produces a significant influence on attachment relationships. Moreover, the effect of CSA on attachment is controlled almost entirely by the form of coping an individual uses under stress. Abused men in this sample who used avoidant or sexualized coping when under stress were likely to have poorer attachment 102 relationships with mother and father. Men with childhood sexual experiences who used non-sexual coping behaviors were likely to have more positive attachment relationships with mother and father. Earlier models indicate that peer relationships are also significant in explaining some of the variability in these relationships, but eliminating them from the model provided a better representation of the data. Childhood sexual experiences did not exert control in current attachment relationships beyond the effect of the coping behaviors used. Exploring specific predictors of sexualized coping and abuse. Although this structural model draws relationships between the main variables of interest, it is difficult to isolate the specific predictors of specific behaviors. To that end, a series of linear regressions were developed based upon the attachment, coping, and abuse indicators identified in models 1 through 9, in addition to adjustment indicators to be explored in hypothesis four. The sample was restricted to participants in Cluster 2 (mean age 21, n = 151) as in previous models. The dependent variables were tested individually, including 30-day masturbation rate and CUSI scores (sex with adults, forced adult sex, and sex with children). In addition to these sexual variables, regressions were developed to predict CSA severity, exposure to multiple forms of abuse, and overall adjustment scores. The attachment and coping predictor variables were entered into the regression in stepwise manner and included four attachment categories provided by the ECL (secure, dismissive, fearful, and preoccupied); mother, father, and peer security; and the five dispositional coping factors (problem-focused, support-seeking, avoidance, acceptance, and self-blame). A summary of the regression output is provided in Table 22. 103 The first series of regressions representing all Cluster 2 members found the 30- day masturbation rate was predicted by insecure maternal attachment. CUSI sex with adults was predicted by dispositional acceptance, CUSI forced adult sex was predicted by dispositional avoidance, and sex with children had no significant predictors. CSA severity was predicted by insecure paternal attachment. Experiencing multiple forms of abuse was predicted by reduced dispositional acceptance and support-seeking, and increased avoidance. The attachment and coping indicators were also used to predict adjustment (Personality Assessment Screener total scores; PAS). Relevant predictors of problematic PAS scores included ECL insecurity, insecure paternal attachment, dispositional avoidance and self-blame, and reduced acceptance coping. These attachment and coping variables accounted for 40% of the variability in PAS adjustment scores. Table 22. Predictors of Sexual Coping, Childhood Abuse, and Adjustment (n = 151) Model Source Predictor R 2 F Std. Coeff. t 30-day Masturbation Mother Security .029 4.340 -.172 -2.083* CUSI SWA D5Accept .040 6.218 .201 2.494* CUSI FAS D5Avoid .054 8.374 .231 2.894** CSA Severity Father Security .043 6.682 -.207 -2.585* CSE Multi D5Accept -.235 -2.876** D5Avoid .176 2.276* D5Support .143 8.194 -.179 -2.199* PAS Total ECL Secure -.271 -4.022*** Father Security -.213 -3.110** D5Avoid .258 3.665*** D5Self Blame .225 3.139** D5Accept .402 19.346 -.150 -2.231* Note. High PAS Scores indicate adjustment problems. Std. Coeff. = Standardized Coefficient; D5Accept=Dispositional Acceptance; D5Avoid=Dispositional Avoidance; D5Self Blame=Dispositional Self-blame; CUSI SWA=Coping Using Sex Inventory Sex With Adults; CUSI FAS=Coping Using Sex Inventory Forced Adult Sex; CSE Multi=multiple forms of abusive childhood sexual experiences. * p<.05; ** p<.01; *** p<.001 104 Using categorical variables, similar logistic regressions were developed to determine significant predictors of Sex Coping and CSA groups. Dependent variables included compulsive masturbation, masturbation under stress, using pressure or force for sex, and CSA group membership (self-identified, family, non-family, perpetration, and none). Predictor variables were the same attachment and coping variables. The results of the logistic regressions showed pressuring someone for sex was predicted by dispositional avoidance, and forced sex was predicted by insecure paternal attachment. Examining CSA categories, membership in the CSA composite group was predicted by insecure paternal attachment, restricted dispositional acceptance, and increased dispositional avoidance. CSA self-identification was predicted by restricted dispositional acceptance. Experiencing familial child sexual abuse was predicted by preoccupied attachment, dispositional avoidance, and reduced dispositional acceptance, while non-familial abuse was predicted by fearful attachment and reduced support- seeking behaviors. Inclusion in the group that had no sexually abusive experiences was predicted by secure paternal attachment, dispositional acceptance, and reduced dispositional avoidance. Summary results are provided in Table 23. 105 Table 23. Predictors of Sex Coping and CSA Groups (n = 151) Model Source Predictor ? 2 ?R 2 S.E Wald df Sig Exp(B) % Correct Pressure D5Avoid 5.193* .07 .048 5.323 1 .021 1.117 89.4 Force Father Security 6.642** .36 1.249 5.1491 1 .023 .059 80.8 CSA Comp D5Avoid .043 4.530 1 .033 1.095 D5Accept .061 8.530 .003 .836 Father Security 21.01*** .22 .483 5.756 1 .016 .314 84.1 CSA SID D5Accept 7.55** .13 .092 7.623 1 .006 .777 94.0 CSA Family D5Avoid .056 6.007 1 .014 1.148 D5Accept .086 7.342 .007 .792 Preoccupied 14.37** .21 .670 4.814 1 .028 4.350 92.1 CSA Non- Family D5Support .077 10.255 1 .001 .782 Fearful 18.81*** .28 .670 5.505 1 .019 4.819 92.7 CSE None D5Avoid .041 4.883 1 .028 .913 D5Accept .059 6.924 1 .009 1.168 Father Security 19.44*** .19 .463 5.621 1 .018 2.994 82.8 Note. D5Avoid=Dispositional Avoidance; D5Accept=Dispositional Acceptance; CSA SID= Self-identified Child Sexual Abuse; CSE None=no abusive childhood sexual experiences. ?R 2 = Nagelkerke R 2 * p<.05; ** p<.01; *** p<.001 The same regressions were run for those in the CSA composite group (n = 27). No attachment or coping variables were statistically significant in predicting current rates of masturbation or CUSI subscale scores (SWA, FAS, SWC). CSA Severity scores were not predicted by any attachment or coping variable in the abused group. Experiencing multiple forms of abuse was predicted by restricted dispositional self blame/venting, and overall adjustment was predicted by ECL security scores. Summary information for the CSA composite group is provided in Table 24. 106 Table 24. Predictors of Abuse and Adjustment among the Abused (n = 27) Model Source Predictor R 2 F Std. Coeff. t CSA Multi D5Self Blame .30 10.865 -.550 -3.296** PAS Total ECL Secure .52 26.882 -.720 -5.185*** Note. Std. Coeff. = Standardized Coefficient; D5Self Blame=Dispositional Self-blame; CSE Multi=multiple forms of abusive childhood sexual experiences. ** p<.01; *** p<.001 Similar logistic regressions were run for variables pertaining to Sex Coping and CSA categories for the abused men. Masturbation Under Stress was predicted by maternal security and reduced problem-focused coping. For men in the CSA composite group, CSA perpetration (sex acts before age 18 with someone 5 or more years younger) was predicted by restricted support-seeking coping behavior. There were no adequate predictors of familial abuse, but diminished support-seeking coping was predictive of non-familial abuse. Results are presented below in Table 25. Table 25. Predictors of Sex Coping and CSA Groups among the Abused Group (n = 27) Model Source Predictor ? 2 ?R 2 S.E Wald df Sig Exp(B) % Correct Masturbation Under Stress D5Problem .108 2.865 1 .091 .833 Mother Security 10.452** .57 1.868 5.815 1 .016 90.398 84.2 CSA Perpetration D5Support 4.675* .32 .164 3.323 1 .068 .741 85.2 CSA Non- Family D5Support 6.885** .30 .109 4.951 1 .026 .784 63.0 Note. D5Problem=Dispositional Problem-focused coping; D5Support=Dispositional Support-seeking. ?R 2 = Nagelkerke R 2 ** p<.01; *** p<.001 107 For comparison, the linear and logistic regressions were run once again with the non-abused men. Abuse categories were omitted from this series of analyses. In the non- abused group (n = 122), a higher 30-day masturbation rate was predicted by maternal insecurity, unlike those in the abused group. CUSI sex with adults was predicted by increased dispositional acceptance and avoidance, while CUSI forced adult sex was predicted by increased dispositional self-blame/venting. CUSI sex with children was predicted in the non-abused group by ECL Preoccupied attachment and dispositional acceptance, while there was no such relationship among the abused men. Poor overall adjustment in the non-abused group was predicted by overall insecurity (ECL), paternal insecurity, and higher levels of dispositional self blame/venting and avoidance. Table 26 provides summary information. Table 26. Predictors of Sex Coping and Adjustment in the Non-abused Group (n = 122) Model Source Predictor R 2 F Std. Coeff. t 30-day Masturbation Mother Security .046 5.542 -.214 -2.354* CUSI SWA D5Accept .204 2.299* D5Avoid .080 5.157 .181 2.038* CUSI FAS D5Self Blame .060 7.559 .244 2.749** CUSI SWC ECL Preoccupied .207 2.339* D5Accept .079 5.049 .209 2.351* PAS ECL Secure -.288 -3.828*** Father Security -.182 -2.384* D5Self Blame .270 3.407** D5Avoid .390 18.510 .219 2.775** Note. Std. Coeff. = Standardized Coefficient; D5Accept=Dispositional Acceptance; D5Avoid=Dispositional Avoidance; D5Self Blame=Dispositional Self-blame; CUSI=Coping Using Sex Inventory; CUSI SWA=Sex with Adults; CUSI FAS=Forced Adult Sex; CUSI SWC=Sex With Children; PAS=Personality Assessment Screener. * p,.05; ** p <.01; *** p<.001 108 Similar logistic regressions were used to predict categorical sex coping variables. Compulsive masturbation (again, this is a period of daily masturbation for a month or more at some previous time) was predicted by increased levels of dispositional self- blame/venting. Masturbation under stress (admitting to being under stress at the time of compulsive masturbation) was predicted by increased dispositional problem-focused coping and insecure peer attachment. Pressuring someone for sex was not related to attachment but to dispositional self-blame/venting. The use of force to gain sex was predicted in the non-abused group by insecure paternal attachment. Results are provided in Table 27. Table 27. Predictors of Sex Coping in the Non-abused Group (n = 122) Model Source Predictor ? 2 ?R 2 S.E Wald df Sig Exp(B) % Correct Compulsive Masturbation D5SelfBlame 5.038* .05 .039 4.721 1 .030 1.088 58.2 Masturbation Under Stress D5Problem .130 7.272 1 .007 1.420 Peer Security 14.427** .38 1.160 5.151 1 .023 .072 89.0 Pressure D5Self Blame 5.487* .10 .072 4.983 1 .026 1.174 91.8 Force Father Security 5.063* .34 1.317 4.231 1 .040 .067 81.8 Note. ?R 2 = Nagelkerke R 2 * p<.05; ** p<.01 Summarizing predictors of coping, abuse, and adjustment. Tables 22 ? 27 provide detailed examination of the independent predictors of coping, abuse, and adjustment. In summary, in accordance with the findings of Smallbone and Dadds (2000, 2001), Tables 22 and 23 show that men who were abused were likely to be insecurely attached to their fathers, used relatively less instrumental 109 coping (support-seeking) and emotion-focused coping (acceptance), and used more avoidance coping than their non-abused counterparts. Men who were insecurely attached to fathers had higher likelihood of experiencing some abusive form of childhood sexual experience, and to have higher CSA severity scores. When examining all participants? attachment and adjustment, attachment security in current relationships (ECL) and paternal security were predictive of overall adjustment. Men with preoccupied attachment (high anxiety, low avoidance) were more likely to have experienced intrafamilial abuse and those fearfully attached (high anxiety, high avoidance) were more likely to have experienced extrafamilial abuse (Table 23). ECL security was the only predictor of adjustment in the abused group (Table 24). There were no significant predictors of recent masturbation rates in the abused group, but earlier reliance on compulsive masturbation to relieve distress was predicted by secure maternal attachment and restricted use of problem-focused coping (Table 25). There were no significant predictors of sexual pressure or force (with peers) among the sexually abused men. However, there was an inverse relationship between abusing someone younger and dispositional support-seeking (instrumental) coping behavior, and since these abused men were less securely attached to fathers, it is reasonable to conclude that paternal attachment may be a factor in the perpetration of abuse (Smallbone & Dadds, 2000, 2001). In accordance with Marshall et al. (2000), within the non-abused group (n = 122), sexual pressure was predicted by high levels of dispositional self-focused, emotional coping (self-distraction, self-blame, and venting) (see Table 27). As found by Ward et al. (1996), sexual force among the non-abused men was predicted by insecure paternal 110 attachment, and fantasies and behavior focused on child sex themes were predicted by preoccupied attachment (high anxiety, low avoidance) in the non-abused group. Paternal security was predictive of being among the non-abused; therefore, although these men report sexual interest in children, it may be that the paternal bond limits acting upon those interests by providing men with a model of appropriate man-to-child relationships. Hypothesis 4 The last hypothesis proposed that coping strategies will predict psychological adjustment and that coping will mediate the relationship between CSA and adjustment. The Personality Assessment Screener (PAS) was used as the measure of adjustment. The PAS provides a total score and ten element scores to represent adjustment. Subscales include Negative Affect (NA), Acting Out (AO), Health Problems (HP), Psychotic Features (PF), Social Withdrawal (SW), Hostile Control (HC), Suicidal Thinking (ST), Alienation (AN), Alcohol Problem (AP), and Anger Control (AC). The PAS total score assesses the potential for clinically significant emotional and behavioral problems and the need for follow-up evaluation (PAS Professional Manual, Morey, 1997). High subscale scores indicate specific areas for potential concern. The screener is not intended as a diagnostic instrument, but as an indicator of problem areas. The PAS subscale scores were entered into the previous model configuration, using the same three sex-coping indicators and avoidant coping linked to the unobserved variable ?Sex Coping? and the remaining four dispositional coping variables linked to the unobserved variable ?Non-sex Coping.? As in previous models, the relationship between the CSA Composite and Sex Coping (not shown) was significant (b = .33, p<.05) as was the relationship between CSA and Non-sex Coping (b = -.26, p<.01). The relationship 111 between the CSA Composite and Adjustment variables was also significant (b = .32, p<.01, R 2 = .10). The direct effect of Adjustment on Sex Coping was b = .72 (p<.01, R 2 = .52), and on Non-sex Coping was b = -.30 (p<.05, R 2 = .09). The partial model is presented below (Figure 18, Model 10) in which all PAS subscales except Hostile Control (HC) are significant. While the model helps establish the significant relationships between variables, fit indices are not yet adequate. Figure 18. Model 10 ? CSA, Coping, and Adjustment .10 Adjust Problems .09 Non-Sex Coping .29 NA e2 .04 AO e3 .33 HP e4 .57 .17 PF e5 .41 .13 SW e6 .36 .01 HC e7 .48 ST e8 .69 .24 AN e9 .08 AP e10 .50 AC e11 .54 D5P e19 .74 .26 D5Ave18 .54 D5S e17 .74 .07 D5SB e16 .21 D5Ac e15 .46 .14 M30e14 .17 Pre13 .23 FASe12 Chi-Sq = 392.207 df = 150 Chi-Sq/df = 2.615 p = .000 GFI = .789 Adjusted GFI = .733 NFI = .481 RMSEA = .104 p = .000 e20 .52 Sex Coping .51 .48 .41 .37 e21 CSA_COMP e22 .26 .49 .32 .72 -.30 p=.00 p=.00 p=.01 .21 -.09 .27 .54 .71 This basic representation indicates poor adjustment accounts for 52% of the variability in Sex Coping, but only 9% of the variability in Non-sex Coping. The NA = Negative Affect AO = Acting Out HP = Health Problems PF = Psychotic Features SW = Social Withdrawal HC = Hostile Control ST = Suicidal Thinking AN = Alienation AP = Alcohol Problems AC = Anger Control 112 composite CSA variable accounts for 10% of the variability in adjustment scores when using all PAS subscales. When all variables are linked to test mediation, the following model is produced (Figure 19, Model 11). Four indicator variables were removed to improve the model?s ability to represent the observed correlation matrix. Figure 19. Model 11 ? Mediation: CSA, Coping, and Adjustment .06 Adjust Problems .12 Non-Sex Coping .16 NA e2 .32 HP e4 .56 .19 PF e5 .43 .09 SW e6 .30 .69 ST e8 .83 .13 AN e9 .50 AC e11 .52 D5P e19 .72 .24 D5Ave18 .55 D5S e17 .74 .21 D5Ac e15 .46 .13 M30e14 .19 Pre13 .25 FASe12 Chi-Sq = 110.841 df = 80 Chi-Sq/df = 1.386 p = .013 GFI = .917 Adjusted GFI = .875 NFI = .788 RMSEA = .051 p = .461 e20 .49 Sex Coping .49 .50 .43 .36 e21 CSA_COMP e22 .36 .24 .64 -.21 p=.02 p=.01 (NS) .40 .71 .18 -.22 (NS) p=.02 .29 .21 .34 -.48 .28 .24 Hypothesis 4 did not include this test of mediation (the effect of adjustment on coping above and beyond the effect of childhood abuse). However, the model provides 113 some useful information. Adjustment mediates the effect of childhood sexual abuse on Sex Coping (correlation is reduced to b = .18, p = .116) but does not mediate the effect of CSA on non-sex coping (correlation reduced to b = -.22, p<.05). Furthermore, the relationship between adjustment and coping (on the right) is reduced from a significant value to non-significance. This model suggests that for those with childhood sexual experiences, high PAS scores (indicating significant adjustment problems) are strong predictors of high rates of masturbation, fantasies and behavior related to forced adult sex, pressuring someone for sex, and avoidant coping behaviors. However, CSA exerts more control than adjustment (high PAS scores) in predicting less non-sexual coping, including problem-focused coping, support-seeking, and acceptance. The best model of the relationships between these variables is provided below (Figure 20, Model 12). The links between coping and adjustment were reversed to demonstrate the impact either form of coping has on current adjustment. Indicators of fit suggest this model adequately reproduces the observed correlations between the CSA, coping, and adjustment variables. 114 Figure 20. Model 12 ? CSA, Coping, and Adjustment .52 Adjust Problems .08 Non-Sex Coping .16 NA e2 .32 HP e4 .57 .19 PF e5 .43 .09 SW e6 .30 .68 ST e8 .82 .13 AN e9 .50 AC e11 .50 D5P e19 .71 .23 D5Ave18 .56 D5S e17 .75 .22 D5Ac e15 .47 .13 M30e14 .19 Pre13 .26 FASe12 Chi-Sq = 109.302 df = 80 Chi-Sq/df = 1.366 p = .016 GFI = .917 Adjusted GFI = .876 NFI = .790 RMSEA = .049 p = .497 e20 .11 Sex Coping .48 .51 .44 .36 e21 CSA_COMP e22 .36 p=.01 p=.04 .33 -.28 p=.02 p=.01 -.23 .29 .21 .34 -.47 .28 .24 .40 .71 -.05 .68 (NS) Interpreting Model 12, both Sexual Coping and Non-sexual Coping mediate the direct effect of CSA on Adjustment. Childhood sexual experiences increase the likelihood of using sexualized coping and avoidance strategies when under stress. These strategies significantly impact adjustment. The presence of childhood sexual experiences also has a negative impact on the use of problem-solving, support-seeking, and acceptance when under stress. In turn, low scores related to non-sex coping behaviors are 115 related to high scores on the PAS, and in this case are particularly predictive of negative affect, health problems, psychotic features (thought disturbances), social withdrawal, suicidal thinking, alienation, and problems with anger control. Based upon the standardized coefficients, it appears that the effect of sexualized coping and avoidance have a stronger impact on adjustment than the absence of adaptive coping. Consolidation of the Models The complex relationships between the variables have been represented in models that highlight both direct and indirect influences between child sexual abuse, coping, attachment, and adjustment. There also have been indications of mediation of some of these direct influences by coping, attachment, and adjustment. To gain a broader conceptualization of the relationships between these four latent variables, they are provided below, free of the indicators variables that have been identified. As a reminder, the most important attachment indicators for this sample were anxiety, mother communication, mother trust, mother alienation, father communication, father trust, and father alienation. The most important indicators of adjustment in this sample were negative affect, health problems, psychotic features, social withdrawal, suicidal thinking, alienation, and anger control. Among the most potent sexualized coping variables were current 30-day masturbation rates, pressuring someone for sex, fantasizing or acting out forced adult sex, and the use of dispositional avoidance coping. The remaining indicators on the non-sexual coping side of the model included problem-focused, support-seeking, and acceptance coping strategies. Figure 21 (Model 13) consolidates the mediational effects of attachment and adjustment while Figure 22 (Model 14) demonstrates the combined effects of coping on current attachment and adjustment. 116 Figure 21. Model 13 ? Mediational Influence of Attachment and Adjustment .04 Attach Problems .47 Adjus t Problems .22 Non-sex Coping .58 Sex Coping CSA_COMP e24e23 e26 e25 -.20 Bold = Significant at p<.05 or better Italic = Non-significant .19 .17 .63 -.26 -.35 .14 .86 Chi-Sq = 291.285 df = 185 Chi-Sq/df = 1.575 p = .000 GFI = .858 Adjusted GFI = .805 NFI = .790 RMSEA = .062 p = .078 -.05 One can examine any three variables to establish mediation. For example, for the Attachment ? Adjustment ? Non-sex Coping triad, it appears that attachment problems mediate the direct effect of adjustment problems on the use of non-sexual coping. Similarly, in the Attachment ? Adjustment ? Sex Coping triad, adjustment problems appear to mediate the direct effect of attachment on the use of sexualized coping. Continuing to examine other triads, attachment problems do not mediate the direct influence of CSA on the use of non-sexual coping, but adjustment problems do increase the use of sexualized coping beyond the influence of CSA. 117 While tests of mediation were the central focus of this research, an unexpected benefit of these models was that reversing directionality provided a demonstration of the effect of sexualized and non-sexualized coping on current attachment and adjustment (Model 14). Figure 22. Model 14 ? Direct Effects of Coping on Current Attachment and Adjustment .27 Attach Problems .82 Adjust Problems .08 Non-sex Coping .11 Sex Coping CSA_COMP e24e23 e26 e25 .33 -.29 Bold = Significant at p<.05 or better Italic = Non-significant .00 .66 .38 -.04 .32 -.40 -.07 -.18 Chi-Sq = 263.013 df = 180 Chi-Sq/df = 1.461 p = .000 GFI = .869 Adjusted GFI = .816 NFI = .811 RMSEA = .055 p = .262 Model 14 consolidates all of the key relationships between childhood sexual abuse, coping styles, and the resulting attachment and adjustment problems that may occur. First and foremost, sexualized coping and non-sexualized coping, as formulated in this research, are orthogonal constructs. Both exert independent effects on attachment and 118 adjustment to the degree that the direct effects of abuse become non-significant. Attachment problems and adjustment problems in this sample of non-offending men are controlled by the type of coping skills they use when under stress. Abused men are more likely to rely on sexualized coping and avoidance and have restricted use of more adaptive coping skills. Subsequently, these men have increased problems with current attachment relationships and psychological adjustment. For those men who use non- sexualized coping, there would be fewer attachment problems, relative to their abused counterparts. Non-sexualized coping also reduces the degree of adjustment problems, but operates through the enhancement of interpersonal relationships that support well-being, rather than directly impacting adjustment. In this study, current attachment problems are controlled by sexualized and non-sexualized coping, while adjustment problems seem to be controlled by sexualized coping and attachment. The combined effects of sex coping and attachment on adjustment accounts for 82% of the variability in adjustment scores, while the combined effects of sex coping and non-sex coping on attachment account for 27% of the variability in attachment scores. Finally, the direct effect of childhood sexual abuse, while statistically significant, accounts for relatively little variability in sexualized coping (11%) and non-sexualized coping (8%), suggesting there are other important components controlling these behaviors. As demonstrated in Model 13, there may some important feedback loops between coping, adjustment, and attachment where one impacts another and produces enhanced or impaired self-efficacy beliefs and self-concepts that persist in future responding (Bandura, 1997). 119 DISCUSSION Overview The primary purpose of this research was to replicate and extend previous work (Lyle, 2003) that suggested a significant relationship between attachment and sexualized coping in non-offending males. In the 2003 study, childhood sexual abuse was significantly related to sexualized coping, and attachment relationships mediated the effect of CSA on sex coping. Additionally, there was some preliminary indication that child sexual abuse directly influenced psychological adjustment, and that quality of attachment mediated that relationship to a modest degree. Findings such as these were indicated by relatively recent models of the impact of childhood sexual abuse on producing sexual offending. For example, Marshall and Marshall (2000) produced an elegant model that suggests poor childhood attachment produces an environment that makes a young boy vulnerable to abuse and limits his ability to cope with the aftermath of abuse. As a result, the boy may develop sexually focused behavior that serves to ease his distress, but also become behaviorally conditioned as a preferred coping strategy. If certain cognitive distortions arise that orient the boy to deviant sexual themes, these themes become reinforced by sexual pleasure, and simultaneously, relief from distress negatively reinforces masturbation. Given some disinhibiting event, this sexually preoccupied adolescent or young man becomes more likely to offend sexually. Smallbone and McCabe (2003) also found support for the 120 relationship between childhood attachment, childhood sexual abuse, early masturbation, and later sexual offending. Smallbone and Dadds (1998, 2000), Marshall and Marshall (2000) and Cortoni and Marshall (2000) provided ample evidence in their research with sex-offenders that poor attachment and sexualized coping were important contributors to the development of sexual offending. In particular, these researchers suggested it was insecure paternal attachment that exerted the most influence in offending. However, the paucity of research into the effects of childhood sexual abuse for men who do not become offenders drove this current project. It became apparent in early literature searches that non-offending sexually abused boys are relatively invisible in the scientific community. At a glance, it appears that less than one percent of the scientific literature devoted to childhood sexual abuse research pertains to male victims, with the predominant focus on those who become offenders. Several researchers cited in this paper have acknowledged that the models of offending behavior are incomplete without examining similar relationships in non-offender samples. The current project incorporated attachment theory (Bowlby, 1988; Alexander, 1992, Berk, 1998; and Thomas, 2000), coping theory (Lazarus & Folkman, 1984; Carver et al. 1989; Bandura, 1997; and Tamres et al., 2002), and sexualized coping research (Duncan & Williams, 1998; Hall et al. 1998; Gartner, 1999; Marshall & Marshall, 2000; Smallbone & McCabe, 2003; and Cortoni & Marshall, 2001) with child sexual abuse research (Finkelhor 1979, 1981; Finkelhor & Browne, 1985; Fromuth & Burkhart, 1987, 1989; Roche et al., 1999). The key variables were examined according to suggested abuse ? effect relationships found in recent research with non-offenders (Rind et al., 1998; 121 Madu & Peltzer, 2001; Merrill, Thomsen, Gold, & Milner, 2001; Smallbone & Dadds, 1998, 2000, 2001; and Stander et al., 2002). This project aimed to better understand the relationships between various types of childhood sexual experiences and the subsequent impact on coping, attachment, and adjustment. Given that sexualized coping had been reported by over half the men in the 2003 college sample, it was anticipated the use of sexualized coping would again be significant in the current project. Moreover, the coping literature suggested that the appearance of sexualized coping might be related to deficits in problem-focused coping. Sex-offender specific models also highlighted the relationship between insecure patterns of attachment and deficits in problem-focused coping. In the broadest terms, men who had poor attachment relationships in childhood failed to develop adequate internal resources to become competent in self-protection, and thereby would be less likely to incorporate problem-solving strategies when under stress. Instead, those with avoidant attachment patterns would tend to use disengagement strategies, while anxiously attached men would resort to more emotion-focused strategies than their securely attached peers. Although sexualized coping has been strongly linked to the development of sexual offending, this finding was not evident when college men reported using similar sexualized coping behavior. In college samples, sexual coping strategies appear to serve a somewhat functional purpose as a measure of self-soothing in the face of poor attachment relationships. Given that so many college men are using this form of coping, regardless of abuse history, it is important to examine differences in typical coping behaviors as they relate to abuse, attachment, and psychological functioning. It may be that the presence of sexualized coping is not sufficient to produce offending, but instead the absence of more 122 appropriate forms of problem-focused coping are potent influences in developing deviant sexual behavior. Furthermore, if the presence of sexualized coping and the absence of more functional coping strategies occur in college samples without producing sexual offending, there are still important components of this phenomenon yet to be revealed. General Findings This college sample was unique in that approximately half of the completed surveys were provided by college men on the campus of a southeastern university, while the other half were provided by men enrolled in various psychology courses across the country. Some participants were solicited through online advertisements to a Psychology Teacher listserv, with interested participants completing the survey online in exchange for extra course credit. The demographic makeup of the sample was examined by format of the survey and significant age differences produced some differences in attachment scores, education levels, parental violence, suicidal ideation and attempts, and dispositional problem-focused coping. The sample age ranged from 19 to 67 years old, therefore most of the exploratory analyses were performed on the full sample and a subset whose mean age was 21 years old. Most of the men were unmarried with no children. The racial makeup of the sample was representative of the college population served at the home university. In comparison with the pilot research for this project, parental involvement was somewhat lower, and number of violent acts committed by father and mother were both higher. Over 20% of the sample came from divorced homes with the mean age at the time of divorce at eight years old. 123 Mental health issues were significant in this sample as well. Over 30% reported suicidal ideation and nearly 6% reported a suicide attempt. Over one fourth of the sample had some contact with mental health professionals. Sexual history of the sample paralleled that found in the pilot research. Ninety-six percent were heterosexual, and the remaining participants were evenly divided across homosexual, bisexual, and unsure groups. Thirteen percent reported a sexual experience contrary to their orientation, and 42 % reported experiencing unwanted sexual arousal. Nearly 60 % of these men reported compulsive masturbation, defined as masturbating every day or nearly every day for a month or more, at some previous point in time. However, only 13 % of the sample acknowledged being under stress at the time they engaged in compulsive masturbation. The average current rate of masturbation was 3 times weekly and 12 times monthly for the period immediately preceding the survey. For those men who experienced some form of childhood sexual abuse, the mean age for onset of masturbation was a year earlier than for the sample (11.65 versus 12.56 years, respectively) and the mean age of first sexual experience was 1 ? years earlier than for the sample (11.45 versus 13.11 years, respectively). Child sexual experiences included family touching, penetration, force, and significant age differences; and non-family touching, penetration, force, and significant age differences. Although only 4 % recognized their childhood sexual experiences as abuse, 7% recognized being sexually abused or forced. When all forms of abusive contact were aggregated, over 19% of the sample had at least one abusive sexual experience. Nearly 4 % of the sample experienced two forms of abuse, 1% experienced three forms, and 1% experienced four forms of abuse. Sexual perpetration before the age 124 of 18 years was admitted by 4.3% of this sample. Mother violence, father violence, suicidal ideation and mental health visits were highly correlated with a number of sexual abuse and sexualized coping variables. Sexualized coping. The Coping Using Sex Inventory indicated these men often used sexual strategies to relieve distress. The predominant focus in this sample pertained to fantasies, masturbation, or behavior directed toward consensual sex, but there was also a slightly higher rate in this sample of fantasies and behavior focused on rape themes. There was relatively little interest in themes related to sex with children. Some preliminary explanations for these CUSI differences, particularly the relative increase in rape themes could be a function of the higher rates of violence these men experienced in the home. Hall et al. (1998) proposed that higher incidences of force by mother and father may predispose men to commit more violent acts in sexual relationships, and entertain more violent fantasies about sexual force. An alternate explanation might be that since approximately 50% of these men were Internet users to some degree, they may have a higher exposure to sexually deviant material as a result of the anonymity provided by the Internet (Forde & Patterson, 1998). Members of the current sample who participated online may have some relative experiences on the Internet that impact a tendency toward violent fantasies; however, there is no way to determine this with the data provided. Attachment. The Experiences in Close Relationships Inventory (ECL) indicated a higher mean Avoidance score and a lower mean Anxiety score for this sample than in the pilot research. Although the ECL served as a modest predictor of frequent masturbation in the 125 earlier project, in the current research the ECL was not an adequate measure by which predictions could be made about compulsive masturbation. When examining current masturbation, higher ECL Anxiety scores were predictive of higher rates of masturbation in the full sample. The Inventory of Parent and Peer Attachment (IPPA) provided scores for mother, father, and peer attachment on dimensions of communication, trust, and alienation. None of the subscale scores was a useful predictor of compulsive masturbation. Mother communication and trust were modest predictors of 30-day masturbation rates, and parent security (combining all parent subscales) adequately predicted 7-day masturbation rates. Coping. Non-sexual coping was measured with the Brief-COPE, which provides fourteen subscales of 2 items each. Five factors were extracted using principal components analysis to allow better representation of coping styles within subsequent models. Situational factors were less stable than dispositional factors, and were excluded from final analyses. While reliability coefficients for the items and subscales were adequate, the reliability coefficients for the extracted situational factors were marginal. Using the dispositional factor structure for analyses, sexualized coping variables were significantly correlated with avoidant coping, and to a lesser degree, self-focused blaming/venting strategies. The avoidant coping factor included scales for behavioral disengagement, denial, and substance use; the self-focused factor included scales for self-distraction, self- blame, and venting. The self-focused blaming factor was modestly correlated with more adaptive forms of coping as well, and was therefore associated in the models with the non-sexual coping variable. 126 Hypothesis one The first hypothesis proposed that attachment patterns would predict specific coping strategies in men who have experienced childhood sexual abuse. Part of this hypothesis rested on replicating previous work (Lyle, 2003) that demonstrated a significant link (standardized coefficient, b = .44, p <.05, R 2 = .19) between attachment and sexualized coping in a sample of non-offending college men. Without that relationship, the remainder of the hypotheses could not stand. It was also expected that differential effects of mother, father, and peer attachment could be determined in models of the attachment ? coping relationships. As expected, the quality of attachment was significantly related to sexualized coping in this sample (b = .39, p <.05, R 2 = .15). Components of this model were somewhat different from the pilot research. In the pilot project, sexualized coping was indicated in the model by the presence of frequent masturbation and the CUSI total score, while in the current project, the indicators of sexualized coping included the 30-day rate of masturbation, the use of pressure to achieve sex, and the CUSI scale representing rape themes. In the previous study, Frequent Masturbation referred to any period of a month or more that the participant masturbated daily, while in the current research, the contemporary 30-day masturbation rate was used in the model. Although Smallbone and Dadds (2001) suggested insecure avoidant attachment was linked to aggression, in this model, attachment avoidance was not significantly related to the pressure and force involved in this sample?s sexualized coping. In contrast, attachment anxiety was a modest predictor of sexualized coping among these college men. The most potent attachment difficulties in this sample were those related to mother 127 alienation (? = .90, p <.001), mother trust (? = -.89, p <.001) and mother communication (? = -.82, p <.001). The remaining significant attachment indicators were (in decreasing strength order) father trust, father communication, and peer trust. The 30-day masturbation rate was the predominant factor in the sexualized coping variable. Once sexualized coping had been established as relevant in this sample, the coping variables were examined in relation to attachment. The Cluster 2 participants (mean age = 21, n = 151) were used for this and subsequent models to help control for age effects in attachment scores. It was immediately apparent (Model 2) that the sexualized coping variables and dispositional avoidance coping were completely unrelated to dispositional problem-focused coping, support-seeking, and acceptance coping strategies, and to a lesser degree, self-focused blaming/venting. This initial observation supported the suggestions made by Merrill et al. (2003), Marshall and Marshall (2000), and Smallbone and Dadds (1998) that sexual behavior becomes a preferred coping strategy by allowing the distressed individual to avoid problems and reduce distress. Developers of the IPPA (Armsden and Greenburg, 1987) have found that secure parental and peer attachment are associated with increased use of problem-focused coping in relation to emotion-focused coping, and that attachment quality in adolescence is associated with less hopelessness, loneliness, and external locus of control; and with greater self-management through coping. In this preliminary coping model, attachment problems were likely to restrict the use of problem-focused, support-seeking, acceptance, and self-focused coping strategies. All attachment variables were significant contributors to coping except IPPA peer communication. 128 The identified coping behaviors appeared to occupy distinctly polar dimensions, with sexualized and avoidant strategies at one end of a continuum and non-sexualized strategies at the other end. In fact, once this model was developed, the influence of attachment on both forms of coping was approximately equal and in opposite directions. Mother attachment was the most potent predictor of both forms of coping behavior with factor loadings often double those of the other attachment variables. Hypothesis one specified certain predictions could be made for men who had abusive childhood sexual experiences; so the variable representing all forms of childhood sexual contact was introduced into the attachment ? coping model. For those men who were abused, there was a significant relationship between abuse and attachment quality (b = .19, p <.05, R 2 = .04), but this relationship accounted for very little variability in attachment relationships. Poor attachment was significantly related to both forms of coping, with a decreased likelihood of using non-sexualized coping (b = -.35, p <.01, R 2 = .12) and an increased likelihood of using sexualized and avoidant coping (b = .38, p<.05, R 2 = .15). The model also suggested that the coping strategies most likely to be impacted by poor attachment (in descending order) would be support-seeking and problem-focused strategies first, followed by acceptance coping and self-focused blaming/venting strategies. Specific predictions were hypothesized for the independent effect of anxious and avoidant attachment; and mother, father, and peer attachment. Linear and logistic regressions were used to examine attachment variables as predictors of sexual and non- sexual coping among those men in the age-restricted subset (Cluster 2, n = 151). Because of the limited number of abused participants, similar equations for the abused group 129 provided little useful information. The work of Alexander (1992), Creasey and Hesson- McInnis (2000), and others suggested that avoidantly attached men experience isolation, dependency and a lack of trust that would reduce their reliance on adaptive coping and increase the use of disengagement and denial when under stress. It was also expected that the abused men would use masturbation to relieve distress. These predictions were not entirely supported. There were no sexualized coping predictors of avoidant attachment, either in Cluster 2 or the abused subset. However, men with high ECL avoidance scores were less likely to seek support (instrumental coping) and more likely to use avoidant coping when distressed. The work of Alexander (1992), and Creasey and Hesson-McInnis (2000) also suggested anxiously attached men would attempt to merge with others through sexual contact and use relatively more emotion-focused strategies. Again, no sexualized coping appeared as a predictor of ECL anxious attachment in Cluster 2 or the abused group. Non-sexual coping strategies were more in accordance with predictions, with increased appearance of self-focused (emotional) strategies including self-blame, self-distraction, and venting. Additionally, there was a reduced likelihood of using support-seeking (instrumental) strategies when distressed. Poor maternal attachment was expected to predict fantasies and behavior focused on forced adult sex among the abused men. No predictors of maternal attachment appeared in the abused subset. In the age-restricted group, poor maternal attachment was predictive of higher current masturbation rates, less fantasy and behavior directed toward consensual sex, less problem-focused coping, and more emotion-focused coping. Restating this in positive terms, securely attached men were less likely to rely on 130 masturbation, more likely to pursue consensual sex, and more likely to use problem- focused strategies when distressed. Examining maternal attachment within the context of ECL attachment groups, men with secure maternal attachment were less likely to use self-focused strategies of self-blame, self-distraction, and venting when distressed. Dismissive attachment produced more current masturbation in relation to insecure maternal attachment. Fearful men and Preoccupied men were less likely to use fantasies and behavior focused on consensual sexual themes as a function of insecure maternal attachment. The maternal attachment relationship seemed important in the regulation of affect through sexual means and problem-solving. In the larger non-abused group, maternal security reduced the likelihood of high rates of masturbation. For these men, their relationships with mothers may have provided them with an internal working model of competence in relationships with women, and engendered behavior directed at consensual sexual relationships. In addition, this interpersonal confidence and security may have impacted problem-focused strategies; related successes improved self-efficacy beliefs (Bandura, 1997) and enhanced further use of these same strategies. Men whose adult attachment styles were fearful or dismissive may have gained some benefit from secure maternal attachment that may have impeded their reliance upon sexual coercion and aggression. Father attachment has been regarded by Marshall and Marshall (2000), and Smallbone and Dadds (1998, 2000, 2001) as the key to understanding the potential for sexually abused boys to become offenders. In line with their findings, it was expected that men with poor paternal attachment would report higher rates of sexual fantasy and 131 behavior directed toward children. However, among the sexually abused men, no sexualized or non-sexualized coping variable was significantly related to paternal attachment. In the larger non-abused group, poor paternal attachment was predictive of increased use of fantasies and behavior related to rape themes, decreased use of support- seeking strategies, and increased use of self-focused blaming/venting strategies. No relationship between paternal attachment and child-focused sexual behavior appeared in this sample. By examining coping at the level of ECL attachment groups, fearful men with poor paternal attachment bonds engaged in relatively more fantasies and behavior related to rape themes. Preoccupied men with poor paternal bonds reported no significant sexualized coping, but were more likely to engage in self-focused blaming, distraction, and venting, and less likely to seek support when distressed. It may be that paternally insecure and fearful men, with high anxiety and avoidance in relationships, have inadequate models for establishing and maintaining loving relationships. Moreover, if paternal violence was modeled in the home these fearful men may affirm their power by channeling their anger and frustration into more violent sexual fantasies. Finally, poor peer attachment was expected to predict more sexual coercion, as these men would find their peer relationships lacking in intimacy. Their attempts to gain secure bonds might be channeled into sexual behavior as a means to achieve closeness (Alexander, 1992). This prediction was partially supported. Dismissive men with poor peer attachment bonds relied more upon using pressure for sex, while fearful men with poor peer attachment relied upon masturbation to relieve distress. Preoccupied men reported less tendency to rely upon sexual behavior and fantasies related to forced sex, and less reliance on support from others. 132 It may be that dismissive men, with high avoidance and low anxiety, relied upon pressure for sexual gratification because they lack the internal resources to develop adequate relationships to produce consensual sexual encounters. Fearful men, with high anxiety and avoidance, may not possess adequate peer relationships to ameliorate their distress. In that context, it appears they turn to sexual release through sexual fantasies and behavior related to rape themes. In summary, tests of the first hypothesis demonstrated there is a significant relationship between abuse and attachment, and attachment quality exerts differential control in many coping behaviors. Sexualized coping and avoidance coping can be modeled as components of dysfunctional coping, and appear to be orthogonal to more adaptive forms of coping. Attachment anxiety was associated with more variation in coping problems. Sexualized coping related to insecure maternal attachment included higher masturbation rates and less consensual sex. Rape themes, and not child sex themes, were related to insecure paternal attachment, particularly among fearful men. Poor peer attachment was expressed by dismissive men through sexual pressure, and by fearful men through masturbation under stress. Overall, the ability to rely on problem- focused coping seemed to rest upon the degree of attachment anxiety and maternal security. Hypothesis two The second hypothesis focused on the effects of childhood sexual abuse on those who self-identified as sexually abused. It was assumed that self-identification was a function of abuse severity (Stander et al., 2002 and Merrill et al., 2003) and would therefore be related to more problematic coping strategies. For those men who self- 133 identified, there was a strong relationship (b = .65, p <.01, R 2 = .45) between abuse and sexualized coping. Sexualized coping in this instance was predominantly related to masturbation under stress, while avoidant coping, forced adult sex, and pressuring someone for sex provide little additional statistical influence. For these men, the recognized sexual exploitation they experienced had expressed its effect through a failure to produce relatively adaptive coping in favor of sexual release. The context of this sexual behavior must be considered when trying to explain this abuse-effect relationship. The variable, Masturbation Under Stress, was provided in response to a two-part question regarding (1) compulsive masturbation at any point in time, and (2) being under stress at the time of such compulsive sexual behavior. It cannot be assumed that these men were currently using this behavior to cope with distress, because there was no significant correlation between 30-day masturbation rates and masturbation under stress for these men. Nonetheless, these men had at some point felt compelled to masturbate daily for a month or more to relieve distress. Although predictions about attachment were not included in the hypothesis, the effect of attachment on these relationships was tested. There was no direct relationship between sexual abuse self-identification and attachment quality. Although attachment problems exerted control in restricting the use of non-sexualized coping, its effect was not a significant influence on the use of sexualized coping for these men. Men who had experienced self-identifiable abuse were less able to construct adaptive coping responses in the face of distress and relied predominantly on self-soothing sexual behaviors. Since attachment was not directly significant in producing sexualized coping, the significant attachment variables must have operated through their effect on non-sexual (adaptive) 134 coping. In this group of abused men, maternal alienation was a strong predictor of difficulties with problem-focused coping, support-seeking, and acceptance. One might conclude that being alienated from mother, and having difficulties with maternal communication and trust (also highly significant for these men) provided them with an internal working model of the self as inadequate and ineffective at generating behaviors that ameliorate distress. A high degree of alienation implies hostility and anger that the child may feel unable to express. Such hostility and anger is theorized to produce sexually aggressive behavior (Smallbone & Dadds, 1998; 2000; 2001), but there has been some disagreement as to the relative importance of maternal versus paternal attachment in sexual offending. In this non-offending sample, maternal alienation seemed to restrict adaptive coping, and in doing so, left the abused boy with few alternatives for regulating distressing emotional states. These men did not rely upon sexual coercion and fantasies/behavior related to forced sex, but instead turned their anger inward upon themselves. Hypothesis three The third hypothesis proposed a test of mediation between child sexual abuse, coping, and attachment. The test of mediation requires that the introduction of a statistically significant third variable reduces a previously significant relationship between the first two variables to zero or non-significance (Baron & Kenny, 1986). In pilot research (Lyle, 2003), child sexual abuse was modestly related to sexualized coping (b = .26, p<.05), and attachment was significantly related to CSA (b = .28, p<.01) and sex coping (b = .37, p<.01). When attachment was introduced as the mediating (third) 135 variable, the direct effect of child sexual abuse on sex coping was reduced to non- significance (b = .16, p = .177). In the present research, the direct effect of childhood sexual abuse on sexualized coping was somewhat stronger (b = .33, p<.05), while the direct effect of attachment on sex coping was equally strong (b = .38, p<.05). The direct effect of abuse on attachment was smaller but significant (b = .19, p<.05), providing the components to test mediation. In the present research, attachment was expected to mediate the abuse ? coping relationship, but did not. The direct effect of abuse on either form of coping was reduced by the direct effect of attachment, but mediation could not be established in this sample. The failure of attachment to mediate coping most logically rests on the different variables used to represent sex coping, and the slightly different measurement of attachment relationships from one study to the next. In the first study (Lyle, 2003), sex coping was indicated by three variables slightly different from the present research, namely Frequent Masturbation, CUSI total score, and pressuring someone for sex. Frequent masturbation referred to any period of daily masturbation lasting 30 days or more, without regard to the recent use of such behavior. The variable was also binary, reported either present or absent at some time. Among college men, the CUSI total score is typically composed largely of scores reflecting consensual sexual themes and less so to behavior directed to rape themes, and so the earlier latent variable, Sex Coping, was composed of somewhat different items. In the present research, the latent variable, Sex Coping, included the participants? current masturbation habits (within the previous 30 days), and included a CUSI subscale related to coercive sexual behavior (fantasizing or acting upon forced sexual relations with others), as well as the similar variable 136 representing pressuring someone for sex. The inclusion of the recent masturbatory patterns made the sex coping variable in this model considerably stronger, and while still significantly related to attachment, this behavior was not controlled entirely by quality of attachment. Paralleling the sex coping side of the model, the non-sexual coping behaviors also were controlled by abuse beyond the direct effect of attachment. Poor attachment restricted the use of problem-focused, support-seeking, and acceptance coping strategies, but not more so than the effect of abuse on the availability of these behaviors. Of interest in the third hypothesis is the reversal of the coping ? attachment links. When CSA predicted coping (sexual and non-sexual), and coping predicted attachment, the effect of coping mediated the direct effect of abuse on attachment, reducing the correlation to nearly zero. This finding suggests that the current attachment relationships of these men are controlled in part by the form of coping they utilize when distressed. Abused men were more likely to rely upon sexualized and avoidant means of dealing with their distressing emotions. For those men who adopt non-sexual coping strategies, the effect is to enhance maternal and paternal attachment, and to a lesser degree, peer attachments. Abused men in this sample who used sexual and avoidant coping strategies when distressed were likely to have poorer current attachments with both father and mother. In Model 9, maternal attachment appeared to be more important to the overall attachment of these men, especially maternal alienation. The effect of sexualized and non-sexualized coping accounted for 33% of the variability in parental attachment scores. Additional linear and logistic regressions were undertaken to dismantle the potent contributors to specific types of abuse and its later expression as sexual force. The 137 findings related to sexualized coping support recent work by Smallbone and Dadds (2000, 2001) that examined sexually coercive behavior in college samples. Although they discovered inconsistencies in the maternal and paternal attachment patterns of college students in relation to coercive sexual behavior, they found preliminary support for parents? loving and rejecting behaviors to influence sexually coercive behavior. Smallbone and McCabe?s (2003) research with incarcerated males found that rapists and intrafamilial child molesters were more likely to report insecure paternal attachment than extrafamilial child molesters, and that offenders with insecure paternal attachment were more likely to report being sexually abused than those with secure paternal attachment. Findings by Smallbone and Dadds (2000, 2001) and Smallbone and McCabe (2003) were partially supported in this research. Within the combined sample of abused and non-abused men, deficits in paternal attachment were associated with vulnerability to sexual abuse, and the use of sexual force upon others. In the combined group, paternal security was also predictive of psychological adjustment. Closer examination of these results indicated that insecure parental attachment was related to the use of sexual force only among the non-abused men, and was a significant predictor of overall adjustment. In the abused group, paternal security was not significantly related to sexualized coping or any abuse category. Adult attachment security as measured by the ECL was predictive of adjustment scores in the abused group. Secure maternal attachment among abused men predicted their using compulsive masturbation at some point to relieve distress. Using masturbation under stress was also predicted by restricted problem-focused coping among abused men. Similarly, non- abused men who reported masturbating under stress showed deficits in problem-focused 138 coping and peer attachment. It is not clear whether the maternal influence on increased masturbation among abused men was a direct effect, or a byproduct of the boy?s transition through abuse. As the developing boy matures, his reliance upon sexual strategies to relieve distress may be transformed by his secure maternal bond, eventually being replaced or diminished by more adaptive forms of coping. Supporting this argument, maternal security among the non-abused predicted lower contemporary masturbation rates, suggesting maternal attachment provided support for developing adaptive coping. Contrary to expectations, perpetration of childhood sexual abuse by men who had been abused was not predicted by any attachment relationship. However, these men reported diminished capacity to use support-seeking coping behavior when distressed. Difficulty with support-seeking was also predictive of being abused by a non-family member. Non-familial abuse was determined by the presence of sexual experiences with a significantly older person, a person of authority (teacher, coach, babysitter, etc.) or by force. Given these restrictions, victims of such non-family sexual experience may have developed difficulty trusting others for instrumental and emotional support. In fact, it is possible that the person who took sexual advantage of the boy or adolescent might have tried to disguise the behavior as ?helping? him (to mature, learn about sex, feel better, etc). In this context it is possible to infer that these boys would find it difficult to ask for help. It cannot be determined from the available data if there is a statistical relationship between type of abuse (at whose hands) and CSA perpetration. In the current model of abuse ? attachment relationships, poor maternal attachment seemed to exert stronger control in the development of masturbatory sexual 139 behavior and avoidant strategies and in the restriction of more adaptive coping. Insecure paternal attachment seemed to be an important contributor to the victimization of these men as children and their later use of sexual force. Hypothesis four The fourth hypothesis linked sexualized and non-sexualized coping to adjustment problems, and suggested that adjustment problems were highly correlated with the use of sexualized coping, and significantly correlated with restricted non-sexualized coping ? particularly problem-focused coping, support-seeking and acceptance. Furthermore, poor adjustment mediated the direct effect of abuse on sexualized coping, but not the direct effect of abuse on non-sexualized coping. The mediational model (Model 11) suggested that abused men may have a restricted range of adaptive coping and an increased tendency to use sexualized coping when under stress. In this model, adjustment problems exert a powerful influence, producing sexualized coping behaviors above and beyond the direct effect of abuse. In contrast, abuse diminishes the availability of non-sexual coping beyond the direct effect of adjustment problems. Restated, the effect of abuse on sexualized coping seems to operate through its effect on adjustment, while abuse restricts adaptive coping in a more direct manner. Reversing the directionality of coping in Model 12 provided additional information. The direct effect of abuse on adjustment was almost entirely mediated by the effect of coping styles on adjustment. Men who used sexual coping behaviors and avoidance were highly likely to demonstrate poor adjustment and men who utilized more adaptive forms of coping were likely to produce better adjustment scores. In this 140 configuration, abuse does not operate directly on adjustment, but instead, adjustment is a product of coping styles and other information not provided in the model. This understanding led to the development of a comprehensive model that demonstrated the interrelationship of the four main variables of interest. Directionality was also explored and supported mediation and direct effects of many variables. The essential features of the comprehensive models (Models 13 and 14) are discussed below. Coping. Coping is impacted by multiple forces as shown in the composite models. Poor coping is a byproduct of abusive experiences, marred by the loss of perceived self- efficacy at ameliorating distress. Coping can be enhanced by the presence of adequate attachment relationships, and current adjustment can hinder or help future coping. At the same time, maladaptive sexualized coping is likely both a cause and effect of attachment problems, and may also exert negative influence on adjustment, creating problems related to emotional control, disrupted thought patterns, alienation, suicidal ideation, and self- control. Sexualized and non-sexualized (adaptive) coping appear to be orthogonal, and exert independent effects and cumulative effects on attachment and adjustment. At the same time, the complex routes by which these opposite coping styles emerge are, in part, determined by attachment history and psychological well-being. Attachment. As modeled, attachment provides a contextual framework for examining coping. It appears that poor attachment relationships impede problem-focused coping and emotion-focused coping, and promote avoidant coping. Men who have been sexually 141 abused are likely to have had particularly troublesome paternal attachment, and may turn to sexual means of self-soothing. Although maternal attachment is highly potent in this sample, it appears that maternal attachment provides the framework for understanding the route of coping behaviors utilized. In this sample, abused men with strong maternal bonds admitted using high rates of masturbatory behavior at some earlier point to relieve distress, but their current rates of masturbation were not relatively high compared to the sample. Secure maternal bonds may have helped these men create more adaptive coping strategies and reduce sexualized strategies. Men with poor maternal bonds were more likely to currently use sex to feel better. Coercive sexualized coping seemed to be controlled by paternal bonds, with a significant relationship between paternal insecurity and use of sexual force. Those men with the poorest paternal bonds, abused or not, demonstrated the highest propensity to use sexual force. Moreover, preoccupied adult attachment in the non-abused group was a significant predictor of sexual interest in children. Limitations Much of this research was exploratory in nature, and relied upon hypotheses developed from current abuse literature pertaining to both men and women victims. While the pilot research gave ample support that these hypotheses were grounded in empirical evidence, this project and its required measurements became a monumental undertaking rather quickly. Gathering information about situational and dispositional coping was an added feature of this research, and using the Brief-COPE as a measure of situational and dispositional coping quickly added to the data system used to draw 142 conclusions. Furthermore, the reliability of the derived dispositional factors was modest at best. In an effort to trim the project, the attachment measures were changed slightly from those used in the pilot research (Lyle, 2003), removing the Parental Bonding Instrument (Parker, Tupling, & Brown, 1979) and updating the Inventory of Parent and Peer Attachment to measure parents independently. The Childhood Sexual Experiences Checklist was revised from the version used by Stander et al. (2002) and again from the version used in pilot research. In making these revisions to the survey instruments, some aspects of the data were more difficult to delineate. For example, the revised CSEC did not allow participants to identify the gender of their abuser, but instead asked if they had an experience contrary to their sexual orientation. The sexualized coping measures were improved by the inclusion of 30-day and 7- day masturbation rates, providing a continuous variable for analyses. However, the dichotomous choice question related to compulsive masturbation was not as useful in this sample. It appeared this past behavior, utilized by nearly 60% of the sample, was not related to any other variable except current masturbation rates. While the pilot research found this to be a potent measure of sexualized coping, in the current project the binary variable was not very informative. In a similar manner, the Coping Using Sex Inventory provided similar results in this research to those of the instrument?s creators (Cortoni & Marshall, 2001), but provided less useful information among non-offenders than was hoped. The majority of men endorsed sexualized coping related to consensual sexual themes, although a slightly higher proportion of participants also were oriented toward rape themes. The CUSI factors used in incarcerated populations might be revised when 143 used in a college sample. There was some indication that a different factor structure could be established. Use of Internet data sampling was both a bonus and a limitation in this project. The availability of more participants made it possible to collect data in a fairly short time period, approximately 90 days. However, 50% of the surveys initiated online were not completed. It is fairly easy to advertise and recruit participants, but it is a complex ordeal to encourage their participation without adequate compensation. Anonymity also posed some challenges, requiring more elaborate login procedures and passwords to allow participants to undertake the long survey in portions as needed. Ultimately, the length and complexity of the survey made it difficult or undesirable for about half of the potential subjects to complete the online survey. Additionally, there were some data glitches in the web design that contributed to several days of lost data. All in all, as a first effort at incorporating online data collection, the inclusion of Internet-based psychology students was an overall success. Along with the different data collection formats came some significant differences that could not be entirely explained. There appeared to be a tendency for those students at the home university to be more securely attached to family and friends and less likely to report suicidal ideation/attempts. Additionally, the two conditions that occurred in a laboratory setting in the presence of the researcher produced somewhat more socially desirable responding. Conversely, in the online formats, across sources, attachment scores were relatively lower, and adjustment problems seemed higher. Many of these differences could be reduced by applying age restrictions to the analyses. However, it was not possible to eliminate these differences without destroying the variability needed 144 to produce abuse ? effect models. Therefore, there are limits to the generalizability of this work. It is encouraging to replicate some of the earlier findings provided from the incarcerated and college samples of other researchers (Smallbone & Dadds, 2000, 2001; Smallbone & McCabe, 2003; Cortoni & Marshall, 2001; Marshall et al., 2000), but it cannot be assumed that these results apply to more diverse populations. Ironically, it may be that future surveys within more diverse populations might be most feasible in an online environment. Conclusions Understanding the sexual abuse of boys has become increasingly important over the past 25 years, beginning with the work of Finkelhor (1981). A large majority of the ensuing research has attempted to identify risk factors for abuse, definitions of the types of sexual contact that may be harmful to a developing boy, and the consequences of such harmful acts. At one time it seemed that sexual abuse prevalence rates were much lower for boys than for girls. In fact, if boys experienced these sexual encounters as pleasurable, it was difficult to justify calling it abuse. However, as investigations became more sophisticated, certain commonalities began to emerge suggesting that early sexualization of boys has psychological, sexual, and social consequences. In contrast, recent meta- analysis (Rind et al., 1998) has argued that sexual abuse produces little detrimental effect beyond that attributed to family environment. This research has attempted to address the predicates of abuse on non-offending men, the relationship between attachment relationships and coping, and the resulting psychological adjustment of abused men compared to non-abused. 145 In the present study, boys experienced inappropriate sexual contact at a rate near 20 percent, although a mere 4 percent identified themselves as sexually abused and 7 percent acknowledged experiencing either abuse or sexual force as a child. Those who self-identified were likely to have experienced a forceful or coercive act that may have been contrary to their sexual orientation, and that happened early enough in their maturation to seem out of place. High rates of masturbation often followed such experiences. It may be argued that the sexualization of young boys presents them with information about their bodies and pleasure that is readily available in times of distress. However, high masturbatory rates were not limited to abused boys. In fact, 60% of this sample engaged in compulsive masturbation at some point in time and 75% of those men denied being distressed at the time. The focus of much of previous male sexual abuse research focused on understanding what compels an adolescent or adult male to transition from victim to offender. In limiting research to populations of incarcerated offenders, investigators were able to dismantle the potential influence of poor attachment bonds and family violence in the appearance of sexual victimization, precocious sexual behavior, sexualized coping, and sexual offending. Yet, this important line of research did less to establish the pathways by which an abused boy successfully processes traumatic sexualization (Finkelhor & Browne, 1985) to produce adaptive coping and ultimate psychological well- being. This research moved toward balancing the scales to begin examining how many boys survive. In doing so, it was necessary to incorporate many of the postulates of sexual offending, and examine how non-offenders compare to offenders in areas of attachment and coping. 146 To that end, this study examined the balance between sexualized coping and adaptive coping, and attempted to define the determinants of coping behaviors for both the abused and non-abused. According to the best models of these relationships, sexualized coping and non-sexualized (adaptive) coping are independent constructs. Maternal attachment plays a significant role in determining the use of adaptive or sexualized coping, but not more so than the direct effect of abuse on coping. Men who have had sexually abusive experiences are less likely to utilize problem-focused and emotion-focused strategies, and rely more upon avoidant and sexualized coping strategies when distressed. This research suggests that indeed, insecure paternal attachment places a boy at higher risk of victimization and later sexual aggression. However, secure maternal attachment may reduce the sexualized effects of this experience by supporting and sustaining adaptive, problem-focused coping. Moreover, it is evident that the form of coping a boy utilizes when he is distressed will in part determine the quality of his current attachment relationships and his overall well-being. Failure of both the maternal and paternal bonds places the boy at highest risk to be abused and to possess few adaptive coping skills beyond sexual means. A natural byproduct of this research was to lend support for the predicates of sexual abuse that have been identified in the sex-offender research literature. The pathways to offending that include early masturbatory behavior and sexual coercion were also found in a small segment of this sample who admitted fantasizing or acting out with children or forcing a peer to engage in sex. The difficulty with strongly endorsing the offender-specific model is that high rates of masturbation and sexualized coping also appear in college samples without producing sexually coercive behavior. The appearance 147 of high rates of masturbation and other forms of sexualized coping are not sufficient to explain or predict sexual coercion. It appears that sexualized coping is an easy alternative during times of distress, but the availability of alternative adaptive coping skills seem to rely upon strong maternal bonds. The current study lends some support to the position of Rind et al., (1998), in that family environment influences the effects of abuse on adjustment. However, the models presented here demonstrate that the effects of child sexual abuse on psychological adjustment may not be directly controlled by family environment but instead are controlled by the coping repertoire available to the victim. These coping behaviors, whether adaptive or avoidant, are regulated in large part by attachment and abuse. The quality of the family environment, and the emotional capital with which boys can develop coping mechanisms, seems to regulate the use of sexual strategies for managing emotional distress. The comprehensive models of abuse-effect relationships presented in this work demonstrate the interplay between abuse, coping, attachment, and adjustment in a non- offending sample of college men. The first comprehensive model outlined that abuse, adjustment, and attachment are all potent contributors to adaptive and maladaptive coping. Adjustment problems will increase the use of sexualized coping beyond the effect of attachment problems, and attachment problems will decrease the use of adaptive coping beyond the effect of maladjustment. There is a direct effect of CSA on adjustment that is not mediated by attachment quality. In a similar manner, there is a direct relationship between CSA and attachment that is not mediated by adjustment. In this 148 configuration, attachment does not mediate the direct effect of abuse on adjustment, but rather controls coping. When examining more causal effects of coping on current adjustment and attachment in the second comprehensive model, it appears that sexualized coping and adaptive, non-sexual coping exert direct and opposite influence on current attachment relationships and overall functioning beyond the direct effect of abuse. Thus, it can be said that attachment problems are not the direct result of abuse but are attributed to poor coping. Similarly, adjustment problems are not directly attributed to abuse but are a result of the combined influence of poor coping and insecure attachment. Summary This research supports previous work (Ward et al., 1996; Smallbone & Dadds, 1998, 2000, 2001; Marshall & Marshall, 2000; Marshall, Serran, & Cortoni, 2000; Cortoni & Marshall, 2001; and Smallbone & McCabe, 2003) suggesting insecure paternal attachment places a boy at higher risk of sexual victimization and later sexual aggression. It appears that strong maternal bonds may reduce the sexual effects of the abuse experience by supporting and sustaining adaptive, problem-focused coping. Moreover, it appears that the form of coping a boy or man utilizes when he is distressed will in part determine the quality of his current attachment relationships and his overall well-being. Although Rind et al. (1998) suggested that the effects of child sexual abuse on current functioning can best be explained by family environment, the evidence in this research did not fully support such conclusions. While family environment may predispose a boy to experience abuse, and may impact his processing of such experiences, the availability of adaptive coping mechanisms is more important in helping 149 regulate the boy?s maturation into a well-functioning adult. Non-offending men with early inappropriate sexual experiences constitute the ?walking wounded? among us. According to the models presented in this work, early sexualization can have detrimental effects on interpersonal relationships, can impede the use of functional coping strategies during times of distress, and can contribute negative valence in the interplay between attachment and psychological adjustment. This research examined the balance between sexualized coping and more adaptive coping as a way to begin understanding how many abused boys arrive at the ?healthier? end of the sexual continuum and seem to live well adjusted lives. There has been a scientific push to understand how abused boys become offenders, and rightly so. It seems though, that an important component of the production of healthier men is the positive counter-influence of adaptive coping in mitigating abuse effects on sexual behavior. While there is a fraction of abused boys who go on to abuse others, many more boys become functional men. Thus, it is vitally important to explore and comprehend how attachment and coping ameliorate the psychological and social effects of male sexual abuse, thereby allowing a large majority of non-offending boys to become non-offending men; and to help explain the full continuum of sexual outcomes from abusive pedophile, through walking wounded, to well-functioning men. 150 REFERENCES Alexander, P. C. (1992). Application of attachment theory to the study of sexual abuse. Journal of Consulting and Clinical Psychology, 60, 185-195. Arbuckle, James L. (1994-1999). AMOS 4.01, SmallWaters Corporation. Chicago, IL. Armsden, G. C., & Greenburg, M. T. (1987). The inventory of parent and peer attachment: Individual differences and their relationship to psychological well-being in adolescence. Journal of Youth and Adolescence, 16, 427-454. Ballon, B. C., Courbasson, C., & Smith, P. D. (2001). Physical and sexual abuse issues among youth with substance abuse problems. Canadian Journal of Psychiatry, 46, 617-621. Bandura, A. (1997). Self-efficacy. The exercise of control. New York: Freeman and Company. Baron, R. A., & Kenny, D. A. (1986). The Moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. Benson, E. (2003). E-research. GradPsych, 1, 26-27. Berk, L. E. (1998). Development through the lifespan. Boston: Allyn & Bacon. Birnbaum, M. (Ed.) (2000). Psychological Experiments on the Internet. San Diego: Academic Press. Summary article available at http://psych.fullerton.edu/mbirnbaum/web/introWeb.htm 151 Birnbaum, M. (2001). Introduction to behavioral research on the Internet. NJ: Prentice Hall. Boman, K. and Bodegard, G. (2000). Long-term coping in childhood cancer survivors: Influence of illness, treatment, and demographic background factors. Acta Paediatr, 89, 105-111. Bowlby, J. (1988). A secure base: Parent child attachment and healthy human development. New York: Basic Books. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measures of adult romantic attachment. An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment Theory and Close Relationships. New York: Guilford. Burkhart, B. R., Fromuth, M. E., Gold, S., & Torquato, S. (1993). Epidemiological analysis of male child sexual abuse. Paper presented at the 1993 meeting of the American Psychological Association, Toronto, Canada. Burt, M., & Katz, B. (1987). Dimensions of recovery from rape: Focus on growth outcomes. Journal of Interpersonal violence, 2, 57-81. Carver, C. S., & Scheier, M. F. (1981). Attention and self-regulation: A control theory approach to human behavior. New York: Springer-Verlag. Carver, C. S., & Scheier, M. F. (1983). A control-theory model of normal behavior, and implications for problems in self-management. In P. C. Kendall (Ed.), Advances in cognitive-behavioral research and therapy, 2, 127-194. New York: Academic Press. Carver, C. S., & Scheier, M. F. (1985). Self-consciousness, expectancies, and the coping process. In Field, McCabe, & Schneiderman (Eds.), Stress and coping, 305-330. Hillsdale, NJ: Erlbaum. 152 Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies. A theoretically based approach. Journal of Personality and Social Psychology, 56, 267-283. Carver, C. S., (1997). You want to measure coping but your protocol?s too long: Consider the Brief COPE. International Journal of Behavioral Medicine, 4, 92-100. Coffey, P., Leitenberg, H., Henning, K., Turner, T., and Bennett, R. (1996). The relation between methods of coping during adulthood with a history of childhood sexual abuse and current psychological adjustment. Journal of Consulting and Clinical Psychology, 64, 1090-1093. Cooper, A., Galbreath, N., and Becker, M. (2004). Sex on the Internet: Furthering our understanding of men with online sexual problems. Psychology of Addictive Behaviors, 18, 223-230. Cooper, A., Scherer, C. R., Boies, S. C., and Gordon, B. L. (1999). Sexuality on the Internet: From sexual exploration to pathological expression. Professional Psychology: Research and Practice, 30, 154-164. Cortoni, F. A., & Marshall, W. L. (2001). Sex as a coping strategy and its relationship to juvenile sexual history and intimacy in sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 13, 27-43. Creasey, G., & Hesson-McInnis, M. (2001). Affective responses, cognitive appraisals, and conflict tactics in late adolescent romantic relationships. Associations with attachment orientations. Journal of Counseling Psychology, 48, 85-96. Duncan, L. E., & Williams, L. M. (1998). Gender-role socialization and male-on-male vs. female-on-male child sexual abuse. A Journal of Research, 39, 765-785. 153 Elmone, P., & Lingg, M. A. (1996). Adult survivors of sexual trauma: A conceptualization for treatment. Journal of Mental Health Counseling, 18, 108-122. Fillion, L., Kovacs, A. H., Gagnon, P., & Endler, N. S. (2002). Validation of the shortened COPE for use with breast cancer patients undergoing radiation therapy. Current Psychology, 21, 17-34. Finch, J. F., Panter, A. T., & Caskie, G. I. L. (1999). Two approaches for identifying shared personality dimensions across methods. Journal of Personality, 67, 407-438. Finkelhor, D. (1979). Sexually victimized children. New York: Free Press. Finkelhor, D. (1981). The sexual abuse of boys. Victimology: An International Journal, 6, 76-84. Finkelhor, D., & Browne, A. (1985). The traumatic impact of child sexual abuse: A conceptualization. American Journal of Orthopsychiatry, 55, 530-541. Folkman, S. and Lazarus, R. (1988). Coping as a mediator of emotion. Journal of Personality and Social Psychology, 54, 466-475. Folkman, S., Lazarus, R. S., Dunkel-Schetter, C., Delongis, A., & Gruen, R. (1986). Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter outcomes. Journal of Personality and Social Psychology, 59, 992-1003. Forde, P., and Patterson, A. (1998). Paedophile Internet Activity. Trends and Issues in Crime and Criminal Justice, 97, 1-6. Frankel, M. and Siang, S. (1999). Ethical and legal aspects of human subjects research on the Internet: A report of a workshop June 10-11, 1999. Accessed at http://www.aaas.org/spp/dspp/sfrl/projects/intres/main.htm on 8/4/03. 154 Fromuth, M. E., & Burkhart, B. R. (1987). Childhood sexual victimization among college men: Definitional and methodological issues. Violence and Victims, 2, 241-253. Fromuth, M. E., & Burkhart, B. R. (1989). Long-term psychological correlates of childhood sexual abuse in two samples of college men. Child Abuse & Neglect, 13, 533-542. Gartner, R. B. (1999). Sexual victimization of boys by men: Meanings and consequences. Journal of Gay & Lesbian Psychotherapy, 3. n.p. Retrieved December, 2001 online. Copies available from Haworth Press, Inc. (1-800-HAWORTH). Hall, D. K., Mathews, F., & Pearce, J. (1998). Factors associated with sexual behavior problems in young sexually abused children. Child Abuse and Neglect, 22, 1045- 1063. Hampton, M. and Frombach, I. (2000). Women?s experience of traumatic stress in cancer treatment. Health Care for Women International, 21, 67-76. Hill, L., & Kennedy, P. (2002). The role of coping strategies in mediating subjective disability in people who have psoriasis. Psychology, Health and Medicine, 7, 261- 269. Hudek-Knezevic, J., Kardum, I., & Vukmirovic, Z. (1999). The structure of coping styles: A comparative study of Croatian Sample. European Journal of Personality, 13, 149-161. Kimerling, R., Ouimette, P., & Wolfe, J. (Eds.) (2002). Gender and PTSD. New York: Guilford Press. Landis, J. T. (1956). Experiences of 500 children with adult sexual deviation. Psychiatric Quarterly Supplement, 30, 91-109. 155 Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer. Lev-Wiesel, R. (1999). Feelings of adult survivors of child abuse toward their offender- parents. Child and Adolescent Social Work Journal, 16, 291-304. Lopez, F. G., & Gormley, B. (2002). Stability and change in adult attachment style over the first-year college transition: Relations to self-confidence, coping, and distress patterns. Journal of Counseling Psychology, 49, 355-364. Lopez, F. G., Mitchell, P., & Gormley, B. (2002). Adult attachment orientations and college student distress: Test of a mediational model. Journal of Counseling Psychology, 49, 460-467. Lyle, P. N. (2003). Adult male survivors of childhood sexual abuse: Attachment and sexualized coping in a non-offender sample. Unpublished master?s thesis, Auburn University, Alabama, U.S. Madu, S. N., & Peltzer, K. (2001). Prevalence and patterns of child sexual abuse and victim-perpetrator relationship among secondary school students in the northern province (South Africa). Archives of Sexual Behavior, 30, 311-321. Marshall, W. L., & Marshall, L. E. (2000). The origins of sexual offending. Trauma, Violence, & Abuse, 1, 250-263. Marshall, W. L., Serran, G. A., & Cortoni, F., A. (2000). Childhood attachments, sexual abuse, and their relationship to adult coping in child molesters. A Journal of Research and Treatment, 12, 17-26. Merrill, L. L., Guimond, J. M., Thomsen, C. J., & Milner, J. S. (2003). Child sexual abuse and number of sexual partners in young women. The role of abuse severity, 156 coping style, and sexual functioning. Journal of Consulting and Clinical Psychology, 71, 987-996. Merrill, L. L., Thomsen, C. J., Gold, S. R., & Milner, J. S. (2001). Childhood abuse and premilitary sexual assault in male Navy recruits. Journal of Consulting and Clinical Psychology, 69, 252-261. 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, 992-1006. Mikulincer, M., Florian, V., & Weller, A. (1993). Attachment styles, coping strategies, and posttraumatic psychological distress. The impact of the Gulf War in Israel. Journal of Personality and Social Psychology, 64, 817-826. Mikulincer, M. (1998). Attachment working models and the sense of trust. An exploration of interaction goals and affect regulation. Journal of Personality and Social Psychology, 74, 1209-1224. Morey, L. C., (1997). Personality Assessment Screener professional manual. Odessa, FL: Psychological Assessment Resources, Inc. Nolen-Hoeksema, S., Parker, L., and Larson, J. (1994). Ruminative coping with depressed mood following loss. Journal of Personality and Social Psychology, 67, 92-104. Parker, G., Tupling, H., & Brown, L. B. (1979). A parental bonding instrument. British Journal of Medical Psychology, 52, 1-10. 157 Pierce, T., & Lyldon, J. (1998). Priming relational schemas. Effects of contextually activated and chronically accessible interpersonal expectations on responses to a stressful event. Journal of Personality and Social Psychology, 75, 1441-1448. Rind, B., Tromovitch, P., & Bauserman, R. (1998). A meta-analytic examination of assumed properties of child sexual abuse using college samples. Psychological Bulletin, 124, 22-53. Roche, D. N., Runtz, M. G., & Hunter, M. A. (1999). Adult attachment: A mediator between child sexual abuse and later psychological adjustment. Journal of Interpersonal Violence, 14, 184-207. Rothbaum, F., Weisz, J. R., & Snyder, S. S. (1982). Changing the world and changing the self: A two-process model of perceived control. Journal of Personality and Social Psychology, 42, 5-37. Scheier, M. F., & Carver, C. S. (1988). A model of behavioral self-regulation: Translating intention into action. In Berkowitz (Ed.), Advances in experimental social psychology, 21, 303-346. New York: Academic Press. Siegel, J. M., Sorenson, S. B., Golding, J. M., Burman, A. M., & Stein, J. A. (1987). The prevalence of childhood sexual assault: The Los Angeles epidemiological catchment area project. American Journal of Epidemiology, 126, 1141-1164. Simoni, J. M., & Ng, M. T. (2000). Trauma, coping, and depression among women with HIV/AIDS in New York City. Aids Care, 12, 567-580. Skinner, J. (1999). Dealing with the aftermath of child sexual abuse: A reflection on three case examples. Pastoral Care, March, 32-36. 158 Smallbone, S. W., & Dadds, M. R. (1998). Childhood attachment and adult attachment in incarcerated adult male sex offenders. Journal of Interpersonal Violence, 13, 555- 573. Smallbone, S.W., & McCabe, B. (2003). Childhood attachment, childhood sexual abuse, and onset of masturbation among adult sexual offenders. Submitted for publication. Smallbone, S. W., & Wortley, R. (2000). Child sexual abuse in Queensland: Offender characteristics and modus operandi. Australian Institute of Criminology, Trends & Issues, 193, 1-6. Stander, V. A., Olson, C. B., & Merrill, L. L. (2002). Self-definition as a survivor of childhood sexual abuse among Navy recruits. Journal of Consulting and Clinical Psychology, 70, 369-377. Swanston, H. Y., Nunn, K. P., Oates, R. K., Tebbutt, J. S., & O?Toole, B. I. (1999). Hoping and coping in young people who have been sexually abused. European Child and Adolescent Psychiatry, 8, 134-142. Tamres, L., Janicki, D., and Helgeson, V. (2002). Sex differences in coping behavior: A meta-analytic review and examination of relative coping. Personality and Social Psychology, 6, 2-30. Thomas, P. M. (2000). Protection, dissociation, and internal roles. Modeling and treating the effects of child abuse. Review of General Psychology, 7, 364-380. Tobin, D. L., Holroyd, K. A., Reynolds, R. V., & Wigal, J. K. (1989). The hierarchical structure of the Coping Strategies Inventory. Cognitive Therapy and Research, 13, 343-361. 159 Whipple, G. M. (1913). Psychology of testimony and support. Psychological Bulletin, 10, 264-268. Wilhelm, K., Roy, K., Mitchell, P., Brownhill, S., & Parker, G. (2002). Gender differences in depression risk and coping factors in a clinical sample. Acta Psychiatrica Scandinavica, 106, 45-53. Wyatt, G. E. (1985). The sexual abuse of Afro-American and white-American women in childhood. Child Abuse & Neglect, 9, 507-519. 160 APPENDICES 161 APPENDIX A SURVEY INSTRUMENTS 162 Experiences in Close Relationships Inventory Instructions for Experiences in Close Relationships Inventory The following statements concern how you feel in romantic relationships. We are interested in how you generally experience relationships, not just what is happening in a current relationship. Respond to each statement by indicating how much you agree or disagree with it. Select the corresponding oval below each statement that most closely matches your feeling about it. Strongly Disagree ? Neutral/Mixed ? Strongly Agree 1-------------2------3------4------5------6-----------7 1. I prefer not to show a partner how I feel. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 2. I worry about being abandoned. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 3. I am very comfortable being close to romantic partners. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 4. I worry a lot about my relationships. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 5. Just when my partner starts to get close to me I find myself pulling away. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 6. I worry that romantic partners won't care about me as much as I care about them. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 7. I get uncomfortable when a romantic partner wants to be very close. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 8. I worry a fair amount about losing my partner. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 9. I don't feel comfortable about opening up to romantic partners. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 10. I often wish that my partner's feelings for me were as strong as my feelings for her/him. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 163 11. I want to get close to my partner, but I keep pulling back. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 12. I often want to merge completely with romantic partners, and this sometimes scares them away. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 13. I am nervous when partners get too close to me. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 14. I worry about being alone. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 15. I feel comfortable sharing my private thoughts and feelings with my partner. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 16. My desire to be very close sometimes scares people away. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 17. I try to avoid getting too close to my partner. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 18. I need a lot of reassurance that I am loved by my partneer. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 19. I find it relatively easy to get close to my partner. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 20. Sometimes I feel that I force my partners to show more feeling, more commitment. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 21. I find it difficult to allow myself to depend on romantic partners. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 22. I do not often worry about being abandoned. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 23. I prefer not to be too close to romantic partners. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 24. If I can't get my partner to show interest in me, I get upset or angry. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 164 25. I tell my partner just about everything. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 26. I find my partner(s) don't want to get as close as I would like. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 27. I usually discuss my problems and concerns with my partner. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 28. When I'm not involved in a relationship, I feel somewhat anxious and insecure. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 29. I feel comfortable depending on romantic partners. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 30. I get frustrated when my partner is not around as much as I would like. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 31. I don't mind asking romantic partners for comfort, advice, or help. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 32. I get frustrated if romantic partners are not available when I need them. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 33. It helps to turn to my romantic partner in times of need. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 34. When romantic partners disapprove of me, I feel really bad about myself. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 35. I turn to my partner for many things, including comfort and reassurance. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 36. I reent it when my partner spends time away from me. Strongly Disagree 0 0 0 0 0 0 0 Strongly Agree 1 2 3 4 5 6 7 Please check that you have answered every statement. Thank You! 165 Inventory of Parent and Peer Attachment Instructions for Inventory of Parent and Peer Attachment This series of statements refers to how you typically relate to your mother, father, and friends. You will see this series of statements three times; once for your mother, once for your father, and once for your friends. Please indicate under each statement how true you believe the statement to be. Remember, you are answering for mother, then father, then friends. You may substitute a mother-figure or father-figure if this applies to you. If you have more than one mother-figure or father-figure, use the relationship that has had the most influence on you. Please rate each statement as best you can. Please rate these statements about your mother (or mother-figure). 1. My mother respects my feelings. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 2. I feel my mother is successful as a parent. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 3. I wish I had a different mother. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 4. My mother accepts me as I am. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 5. I can't rely on my mother when I have a problem to solve. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 6. I like to get my mother's point of view on things I am concerned about. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 7. I feel it's no use letting my feelings show with my mother. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 8. My mother senses when I am upset about something. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 9. Talking over my problems with my mother makes me feel ashamed or foolish. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 10. My mother expects too much from me. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 166 11. I get upset easily at home with my mother. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 12. I get upset a lot more than my mother knows about. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 13. When we discuss things, my mother considers my point of view. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 14. My mother trusts my judgment. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 15. My mother has her own problems, so I don't bother her with mine. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 16. My mother helps me to understand myself better. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 17. I tell my mother about my problems and troubles. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 18. I feel angry with my mother. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 19. I don't get much attention at home from my mother. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 20. My mother encourages me to talk about my difficulties. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 21. My mother understands me. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 22. I don't know if I can depend on my mother these days. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 23. When I am angry about something, my mother tries to be understanding. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 24. I tust my mother. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 25. My mother doesn't understand what I am going through these days. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 167 26. I can count on my mother when I need to get something off my chest. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 27. I feel my mother doesn't understand me. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 28. If my mother knows something is bothering me, she asks me about it. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 Please rate these statements about your relationship with your father (or father- figure). 1. My father respects my feelings. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 2. I feel my father is successful as a parent. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 3. I wish I had a different father. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 4. My father accepts me as I am. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 5. I can't rely on my father when I have a problem to solve. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 6. I like toget my father's point of view on things I am concerned about. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 7. I feel it's no use letting my feelings show with my father. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 8. My father senses when I am upset about something. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 9. Talking over my problems with my father makes me feel ashamed or foolish. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 10. My father expects too much from me. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 168 11. I get upset easily at home with my father. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 12. I get upset a lot more than my father knows about. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 13. When we discuss things, my father considers my point of view. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 14. My father trusts my judgment. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 15. My father has his own problems, so I don't bother him with mine. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 16. My father helps me to understand myself better. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 17. I tell my father about my problems and troubles. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 18. I feel angry with my father. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 19. I don't get much attention at home from my father. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 20. My father encourages me to talk about my difficulties. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 21. My father understands me. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 22. I don't know if I can depend on my father these days. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 23. When I am angry about something, my father tries to be understanding. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 24. I trust my father. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 25. My father doesn't understand what I'm going through these days. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 169 26. I can count on my father when I need to get something off my chest. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 27. I feel my father doesn't understand me. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 28. If my father knows something is bothering me, he asks me about it. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 Please rate these statements about your friends. 1. I like to get my friends' point of view on things I'm concerned about. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 2. My friends sense when I'm upset about something. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 3. When we discuss things, my friends consider my point of view. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 4. Talking over my problems with my friends makes me feel ashamed or foolish. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 5. I wish I had different friends. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 6. My friends understand me. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 7. My friends encourage me to talk about my difficulties. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 8. My friends accept me as I am. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 9. I feel the ned to be in touch with my friends more often. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 10. My friends don't understand what I'm going through these days. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 170 11. I feel alone or apart when I am with my friends. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 12. My friends listen to what I have to say. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 13. I feel my friends are good friends. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 14. My friends are fairly easy to talk to. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 15. When I am angry about something, my friends try to be understanding. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 16. My friends help me to understand myself better. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 17. My friends are concerned about my well-being. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 18. I feel angry with my friends. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 19. I can count on my friends when I need to get something off my chest. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 20. I trust my friends. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 21. My friends respect my feelings. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 22. I get upset a lot more than my friends know about. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 23. It seems as if my friends are irritated with me for no reason. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 24. I tell my friends about my problems and troubles. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 25. If my friends know something is bothering me, they ask me about it. Almost Never True 0 0 0 0 0 Almost Always True 1 2 3 4 5 171 Childhood Sexual Experiences Checklist Part A: Demographics Today's date ___________ Your Birth Date ______________ (mm/dd/yyyy) (mm/dd/yyyy) Your Age ______ Your Gender Select... Male __ Female __ Racial/Ethnic Group Select...African American __ Native Aerican _ Hispanic __ Asian __ Caucasian __ Other __ What is the highest level of education you've completed? Select one... Some High School __ High School Diploma or GED __ Some College/Technical School __ Bachelors Degree __ Graduate School __ Masters Degree or higher __ What is your marital status? Select one... Single __ Maried _ Divorced _ Widowed __ Cohabitaing _ How many children (natural, adopted, or stepchildren) do you have under age 18? _______ CONTINUE ON NEXT PAGE 172 Part B: Family History This section asks about the behavior of your family and friends when you were growing up, and about your emotional experiences. 1. Are your biological parents living together? Yes __ No __ 2. Are your biological parents divorced? Yes __ No __ 3. If your biological parents are divorced, how old were you when they divorced? ______ years old 4. Up to the age of 18, how many years was your mother (or stepmother) living in your home? (0 to 18 years) _________ 5. How involved was your mother in raising you? (Check One) Not at all involved 0 0 0 0 0 Extremely involved 6. Up to the age of 18, how many years was your father (or stepfather) living in your home? (0 to 18 years) _________ 7. How involved was your father in raising you? (Check One) Not at all involved 0 0 0 0 0 Extremely involved 8. Were you ever a foster child? Yes __ No __ 9. How strict were your parents in making you obey their rules? (Check One) Not at all strict 0 0 0 0 0 Extremely strict 173 These next 2 questions about violence in the home refer to acts such as hitting, kicking, throwing someone down, or choking someone. 10. How many times did your father (or father figure) use violence in the home toward a spouse or child? Select one... never __ 1-5 times __ 6-10 times __ 11-15 times __ 16-20 times __ > 20 times __ 11. How many times did your mother (or mother figure) use violence in the home toward a spouse or child? Select one... never __ 1-5 times __ 6-10 times __ 11-15 times __ 16-20 times __ > 20 times __ 12. Did you ever feel uncomfortable hanging around with your friends for fear that you might get in trouble while you were with them? (Check One) Not at all uncomfortable 0 0 0 0 0 Extremely uncomfortable 13. While you were growing up, did any of your friends ever regularly get into trouble with the law (like truancy, fighting, or running away)? (Check One) Not at all 0 0 0 0 0 Frequently 14. While you were growing up, did you ever run away from home for more than 24 hours? Yes __ No __ 15. Have you ever thought of suicide to the point of considering how to commit suicide or kill yourself? Yes __ No __ 16. Have you ever attempted suicide or attempted to kill yourself? Yes __ No __ 17. How many visits have you made to a mental health counselor or therapist? (Choose one) (none) 0 0 0 0 0 0 0 0 0 0 0 (10 or more visits) 0 1 2 3 4 5 6 7 8 9 10 174 Part C: Sexual Experiences This Section deals with sexual material of a sensitive nature. Please respond as accurately as you can. 1. At what age did you begin masturbating? (0-18) ______ years old (zero for NEVER) 2. At what age did you have your first sexual experience with another person? (This includes sexual touching, fondling, oral sex, vaginal or anal intercourse, or any other interpersonal sexual contact, even when you were a child) (0-18) ______ years old (zero for NEVER) 3. Was your first sexual experience with your consent? Yes __ No __ 4. What was your primary sexual orientation at the time of your first sexual experience? Select one... Heterosexual __ Homosexual __ Bisexual __ Unsure __ 5. Did your first experience match your sexual orientation? Yes __ No __ Unsure__ 6. At what age were you first aware of your sexual orientation? (0-18) _______years old (zero for NOT SURE) 7. What is your primary sexual orientation now? Select one... Heterosexual __ Homosexual __ Bisexual __ Unsure __ 8. Have you ever had a sexual experience that did not match your orientation? Yes __ No __ 175 9. Before the age of 13, have you ever been forced into sex of any kind against your will? Yes __ If Yes, at what age(s) did it happen? (Mark all that apply, ages 1-12) No __ 0 0 0 0 0 0 0 0 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 10. Between the age of 13 and 18, have you ever been forced into sex of any kind against your will? Yes __ If Yes, at what age(s) did it happen? (Mark all that apply, ages 13-18) No __ 0 0 0 0 0 0 13 14 15 16 17 18 11. Have you ever been sexually aroused by another person when you did not want to be? Yes __ If Yes, did the arousal bring up uncomfortable emotions for you? No __ Select one... Fear __ Disgust__ Guilt __ Shame __ Sadness__ Anger __ Other __ 12. Have you ever experienced a period of time in which you masturbated every day or nearly every day for a month? Yes __ No __ If Yes, was this period of masturbation related to a stressful time in your life? Yes __ No __ If Yes, briefly describe the stressful event_________________________________ How upset were you at the time you engaged in frequent masturbation (described above)? (Check One) Not at all upset 0 0 0 0 0 Extremely upset 13. How many times in the past week have you masturbated? (number)_______ 14. About how many times in the past month have you masturbated? (number)_______ Were you upset at the time? (Check One) Not at all ___ Very rarely ___ Sometimes ___ Usually ___ Always ___ 176 15. Were you sexually abused as a child (prior to age 13)? Yes __ If Yes, at what age(s) did it happen? (Mark all that apply, ages 1-12) No __ 0 0 0 0 0 0 0 0 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 16. Were you sexually abused as an adolescent (age 13 to 18)? Yes __ If Yes, at what age(s) did it happen? (Mark all that apply, ages 13-18) No __ 0 0 0 0 0 0 13 14 15 16 17 18 If you were sexually abused, think about the most upsetting experience: How upset were you at the time? (Check One) Not at all upset 0 0 0 0 0 Extremely upset How upset are you about it now? (Check One) Not at all upset 0 0 0 0 0 Extremely upset 17. With approximately how many females have you had sexual intercourse, whether voluntary or forced? (By intercourse we mean oral, anal, or vaginal penetration) (0-100) __________ 18. With approximately how many males have you had sexual intercourse, whether voluntary or forced? (By intercourse we mean oral or anal penetration) (0-100) __________ 19. When you have voluntary intercourse, is a condom used? Select one... Never __ Sometimes __ Almost always__ Always __ Please continue 177 Part D: Sexual History This section asks about different types of sexual experiences you have had before the age of 18. Please think carefully about your past experiences and report as many of those experiences as you can remember. These experiences include kissing, sexual touching, and any kind of penetration (oral, anal, or vaginal). 1. Before you were 18 years old, have you had any form of sexual contact with a family member? (if you are unsure, look at the list below to see the types of people we mean) Yes __ No __ Family Members include Check here for YES Cousin, niece/nephew, or distant relative ___ Aunt or Uncle ___ Grandparent ___ Sibling or Step-sibling ___ Parent or Stepparent ___ If Yes, use the following section to record the details. Total number of times it happened _____ Total number of years it happened _____ Total number of different people _____ Any Touching/Kissing? (Yes/No)____ How many times?____ Any Penetration? (Yes/No)____ How many times?____ Were any of these experiences with a person 5 years older than you? (Yes/No)____ How many experiences? ____ Were any of these experiences with a person 5 years younger than you? (Yes/No)____ How many experiences? ____ Did any of these experiences upset you at the time? Yes __ No __ If Yes, Pick the worst time and mark how upsetting it was at the time (Check One) Not at all upset 0 0 0 0 0 Extremely upset Recalling the worst time, how upset are you about it now? (Check One) Not at all upset 0 0 0 0 0 Extremely upset 178 Now, think about sexual contact you had with non-family members 2. Before you were 16 years old, have you had any form of sexual contact with a non- family member? (if you are unsure, look at the list below to see the types of people we mean) Yes __ No __ (I?ve had sex with) Non-family members Check here for YES Boyfriend or Girlfriend ___ Acquaintance or Stranger ___ Babysitter ___ Teacher or Coach ___ Authority figure (Boss, Supervisor) ___ Trusted person (Clergy, Counselor) ___ If Yes, use the following section to record the details. Total number of times it happened _____ Total number r of years it happened _____ Total number of different people _____ Any Touching/Kissing? (Yes/No)____ How many times?____ Any Penetration? (Yes/No)____ How many times?____ Were force or threats ever used? (Yes/No)____ How many times?____ Were any of these experiences with a person 5 years older than you? (Yes/No)____ How many experiences? ____ Were any of these experiences with a person 5 years younger than you? (Yes/No)____ How many experiences? ____ Did any of these experiences upset you at the time? Yes __ No __ If Yes, Pick the worst time and mark how upsetting it was at the time (Check One) Not at all upset 0 0 0 0 0 Extremely upset Recalling the worst time, how upset are you about it now? (Check One) Not at all upset 0 0 0 0 0 Extremely upset 179 These last two questions refer to sex you have initiated with another person. 3. Have you ever used pressure to coerce someone into having sex with you? Yes __ If Yes, how many times? _________ No __ 4. Have you ever used force to get sex from someone? Yes __ If Yes, how many times? _________ No __ Thank-you. You have completed this part of the survey. 180 Coping Using Sex Inventory Instructions for Coping Using Sex Inventory The following statements are ways people react to various difficult, stressful, or upsetting situations. Please indicate how much you engage in those types of activities when you encounter a difficult, stressful, or upsetting situation. Not at all - Infrequently - Sometimes - Often - Almost always 1 2 3 4 5 1. I fantasize about having sex with a consenting adult. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 2. I fantasize about having sex with a child. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 3. I fantasize about forcing an adult to have sex. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 4. I have sex with my regular partner. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 5. I have sex with a child. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 6. I go out and "score" with a stranger. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 7. I masturbate while fantasizing about a consenting adult. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 8. I masturbate while fantasizing about raping an adult. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 9. I masturbate while fantasizing about a child. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 10. I masturbate while fantasizing about hurting someone. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 11. I use pornography depicting consenting adults. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 181 12. I use violent pornography. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 13. I use pornography depicting children. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 14. I masturbate while using pornography. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 15. I go out and rape someone. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 16. I force my regular partner to have sex. Not at all 0 0 0 0 0 Almost always 1 2 3 4 5 Please check that you have answered every statement. Thank You! 182 Brief COPE Inventory Instructions for Brief COPE Inventory This Inventory will evaluate how you cope with stressful or upsetting situations. Please mark the oval that best indicates how much each statement applies to you. Please identify the most upsetting situation you have experienced IN THE LAST MONTH (write here)___________________________________________________________________ Please rate how upsetting this situation was at its worst. Not at all 0 0 0 0 0 0 0 0 0 0 Very much so 1 2 3 4 5 6 7 8 9 10 The following statements include many different ways people cope with upsetting situations. As you read each statement, rate how much you used that strategy to cope with the situation you named above. 1. I've been concentrating my efforts on doing something about the situation I'm in. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 2. I've been taking action to make the situation better. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 3. I've been trying to come up with a strategy about what to do. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 4. I've been thinking hard about what steps to take. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 5. I've been trying to see it in a different light to make it seem more positive. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 6. I've been looking for something good in what is happening. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 7. I've been accepting the reality of the fact that it has happened. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 183 8. I've been learning to live with it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 9. I've been making jokes about it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 10. I've been making fun of the situation. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 11. I've been trying to find comfort in my religion or spiritual beliefs. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 12. I've been praying or meditating. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 13. I've been getting emotional support from others. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 14. I've been getting comfort and understanding from someone. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 15. I've been trying to get advice or help from other people about what to do. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 16. I've been getting help and advice from other people. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 17. I've been turning to work or other activities to take my mind off things. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 18. I've been doing something to think about it less, such as going to movies, watching TV, reading, daydreaming, sleeping, or shopping. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 19. I've been saying to myself "this isn't real." Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 20. I've been refusing to believe that it has happened. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 21. I've been saying things to let my unpleasant feelings escape. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 184 22. I've been expressing my negative feelings. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 23. I've been using alcohol or other drugs to make myself feel better. Not at all 0 0 0 0 0 Very much so 24. I've been using alcohol or other drugs to help me get through it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 25. I've been giving up trying to deal with it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 26. I've been giving up the attempt to cope. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 27. I've been criticizing myself. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 28. I've been blaming myself for things that happened. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 Please rate how much you are still upset since trying to cope with the situation this past month. Not at all 0 0 0 0 0 0 0 0 0 0 Very much so 1 2 3 4 5 6 7 8 9 10 ------------------------------------------------------------------------------------------------------------- Now, please rate how you TYPICALLY respond to upsetting events. This is a general impression you have of the way you usually handle things. 1. I concentrate my efforts on doing something about the situation. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 2. I take action to try to make the situation better. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 3. I try to come up with a strategy about what to do. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 4. I think hard about what steps to take. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 185 5. I try to see it in a different light to make it seem more positive. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 6. I look for something good in what is happening. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 7. I accept the reality of the fact that it has happened. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 8. I learn to live with it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 9. I make jokes about it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 10. I make fun of the situation. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 11. I try to find comfort in my religion or spiritual beliefs. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 12. I pray or meditate. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 13. I get emotional support from others. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 14. I get comfort and understanding from someone. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 15. I try to get advice or help from other people about what to do. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 16. I get help and advice from other people. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 17. I turn to work or other activities to take my mind off things. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 18. I do something to think about it less, such as going to movies, watching TV, reading, daydreaming, sleeping, or shopping. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 19. I say to myself "this isn't real." Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 186 20. I refuse to believe it happened. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 21. I say things to let my unpleasant feelings escape. Not at all 0 0 0 0 0 Very much so 22. I express my negative feelings. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 23. I use alcohol or other drugs to make myself feel better. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 24. I use alcohol or other drugs to help me get through it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 25. I give up trying to deal with it. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 26. I give up the attempt to cope. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 27. I criticize myself. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 28. I blame myself for things that happened. Not at all 0 0 0 0 0 Very much so 1 2 3 4 5 Please check that you have answered every item. Thank You! 187 Personality Assessment Screener Morey, L. C. (1997) (copyright material not presented here) 188 APPENDIX B INFORMATION LETTER AND RESOURCES 189 INFORMATION LETTER FOR ---The Role of Attachment and Coping in Long-term Psychological Adjustment--- You are invited to participate in a research study examining the relationship between patterns of attachment and coping behavior in determining responses to stress. This study is being conducted by Patricia Lyle, M.S., clinical psychology graduate student, under the supervision of Barry Burkhart, Ph.D. We hope to learn if there is a significant relationship between attachment style and the way men behave when they are under stress. You must be male and at least 19 years of age to participate. If you decide to participate, we will ask you to complete a comprehensive questionnaire, which may take approximately one (1) hour of your time. Auburn students will receive two (2) hours of extra credit for participating. No identifying information will be collected. Therefore, your responses on this questionnaire are completely anonymous. This anonymous questionnaire covers a wide range of information about your attachment to your family and friends, your coping skills when under stress, your current psychological health, and your sexual history. You will be asked about a wide variety of sexual behaviors that you might have engaged in including masturbation, oral sex, anal sex, sexual touching, kissing, and more. Not everyone will have such experiences to report, and not everyone will feel comfortable revealing these types of activities. There will also be questions about your sexual orientation and possible sexual experiences you may have had during your childhood. Again, some people may not feel comfortable revealing this information. Some people may believe they have no important sexual experiences to report, but even those with very little sexual experience are eligible to participate. This study will examine many types of male sexual behavior, and many levels of sexual experience. Some men may be embarrassed by the sensitive nature of the questions. If you believe answering questions about your current or past sexual experiences will cause you distress, please do not participate in this study. You will be compensated with two (2) hours of extra credit for your participation if you are currently enrolled in Auburn University Psychology courses. You will not receive extra credit if you are not currently enrolled in Auburn University Psychology courses. Additionally, the results of this research may add to the scientific knowledge regarding men?s early sexual experiences, and the role of attachment in their coping behavior, especially during times of stress. HUMAN SUBJECTS OFFICE OF RESEARCH PROJECT #04-128 AR 0409 APPROVED 9-14-04 TO 9-13-05 190 Any information obtained in connection with this study will remain anonymous. Data will be collected regarding your age, level of education, marital status, and number of children, but no specific identifying information will be collected. Information collected through your participation may be used to fulfill an educational requirement for a doctoral dissertation, may be published in a professional journal, and/or presented at a professional meeting, etc. You may withdraw from participation at any time, without penalty. However, due to anonymity, your data cannot be isolated or retrieved from the set. Your decision whether or not to participate will not jeopardize your future relations with Auburn University or the Psychology department. If you have any questions we invite you to ask them now. If you have questions later, Patricia Lyle (lylepan@auburn.edu, 844-4932) will be happy to answer them. You may also contact Dr. Burkhart, Ph.D. (burkhbr@auburn.edu, 844-6476) if needed. Any contact with either person by phone or email will be treated as confidential. You will be provided a copy of this form to keep. A referral list is also provided with this information letter if you would like to seek outside counseling as a result (direct or indirect) of your participation in this project. For more information regarding your rights as a research participant you may contact the Office of Research Programs by phone or e-mail. The people to contact there are Mr. Chip Burson, Executive Director at (334) 844-5966 (bursoen@auburn.edu) or Dr. Peter Grandjean at (334) 844-1462 (grandpw@auburn.edu). HAVING READ THE INFORMATION PROVIDED, YOU MUST DECIDE WHETHER OR NOT YOU WISH TO PARTICIPATE IN THIS RESEARCH STUDY. ___________________________________ Investigator's signature Date HUMAN SUBJECTS OFFICE OF RESEARCH PROJECT #04-128 AR 0409 APPROVED 9-14-04 TO 9-13-05 191 REFERRAL LIST OF AUBURN-AREA MENTAL HEALTH SERVICE PROVIDERS Individual/Agency Services Available Cost/Hour East Alabama Mental Health Center Individual and group therapy $8-80 (334)742-2700 Based on (334) 821-0660 (After hours emergency #) income Student Counseling Services Individual and group therapy No charge Auburn University (334) 844-5123 Auburn Univ. Psychological Services Marriage, family, and $25 -55 (334) 844-4889 individual therapy Based on incom Clinical Psychologists Individual and group therapy $75-100 248 E. Glenn Ave. (334) 821-3350 Anne Harzem Marriage, family, and $90 2204 Executive Park Dr., Opelika individual therapy (334) 745-0923 Nana Daranasty Individual and group therapy $30-75 318 N. College St Based on (334) 821-9770 income Crisis Center Phone counseling No charge (334) 821-8600 Rape Counselors of East Alabama Phone counseling No charge (334) 745-8634