|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 002006002
007 cr mnu|||uuuuu
008 090605s2008 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002708
The relationship between childhood and adolescent family environment and adult psychological functioning in females who experienced childhood sexual abuse
h [electronic resource] /
by Ross Krawczyk.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 40 pages.
Thesis (M.A.)--University of South Florida, 2008.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: Research has shown childhood sexual abuse (CSA) to be related to many negative outcomes in adulthood including psychopathology. Findings in this area, however, are very inconsistent, with the relationship between CSA and adult outcomes varying greatly across studies. This relationship is further complicated by the co-occurrence with CSA of other risk factors in childhood. The present study examines the prediction of adult psychopathology, measured by the Brief Symptom Inventory (BSI; Derogatis, 1982), made by CSA, measured by the Early Sexual Experiences Survey (ESE; Bartoi & Kinder, 1998), childhood SES (Hollingshead, 1975), parental bonding, as measured by the Parental Bonding Instrument (PBI; Parker, Tupling, & Brown, 1979), and parental separation/divorce. It was hypothesized that CSA, SES, PBI, and parental separation/divorce would significantly predict BSI scores. It was also hypothesized that CSA would significantly predict BSI scores beyond the variance accounted for by the other variables. Results indicated that all predictor variables were significantly related to BSI score in the hypothesized direction, except for childhood SES which was found to be unrelated to BSI score in adulthood. A regression model including parental care, overprotection, and divorce/separation significantly predicted BSI score. When objective and subjective CSA severity scores were added to the equation, the amount of variance in BSI score accounted for significantly increased. Amounts of shared variance were quite high, but results indicated that CSA severity accounts for variance in adult psychological functioning beyond that accounted for by parental care, overprotection, and divorce.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Bill N. Kinder, Ph.D.
Childhood sexual abuse
t USF Electronic Theses and Dissertations.
i The Relationship between Childhood and A dolescent Family Environment and Adult Psychological Functioning in Females Who Experienced Childhood Sexual Abuse by Ross Krawczyk A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Bill N. Kinder, Ph.D. Vicky Phares, Ph.D. Marcie Finkelstein, Ph.D. Date of Approval: September 12th, 2008 Keywords: Childhood Sexual Abuse, Psychopathology, Abuse Severity, Parenting Copyright 2008, Ross Krawczyk
ii Table of Contents List of Tables iii Abstract iv Introduction 1 Hypotheses 12 Method 13 Participants 13 Measures 13 Demographics 13 Psychological Functioning 13 Childhood Sexual Abuse 14 Childhood Socioeconomic Status 15 Parental Bonding 15 Procedure 16 Results 17 Descriptive Statistics 17 Hypotheses 18 Discussion 23 Conclusions 27 Implications 28 References 31 Appendices 36 Appendix A: ESE-R 36 Appendix B: Demographic Questi onnaire 37 Appendix C: Parental Bonding Inventory 40
iii List of Tables Table 1 Correlation Matrix: CSA & BSI subscales 19 Table 2 Correlation Matrix: SES, PBI s ubscales, & BSI 20 Table 3 Regression Model: Parenting Variables 21 Table 4 Regression Model: CSA Vari ables 22 Table 5 Regression Model: Combined Model 22
iv The Relationship between Childhood and A dolescent Family Environment and Adult Psychological Functioning in Females Who Experienced Childhood Sexual Abuse Ross Krawczyk ABSTRACT Research has shown childhood sexual abuse (CSA) to be related to many negative outcomes in adulthood including psychopathology Findings in this area, however, are very inconsistent, with the relationship be tween CSA and adult outcomes varying greatly across studies. This relationship is further complicated by the co-occurrence with CSA of other risk factors in childhood. The present study examines the prediction of adult psychopathology, measured by the Brief Symp tom Inventory (BSI; Derogatis, 1982), made by CSA, measured by the Early Sexual Ex periences Survey (ESE; Bartoi & Kinder, 1998), childhood SES (Hollingshead, 1975), parental bonding, as measured by the Parental Bonding Instrument (PBI; Park er, Tupling, & Brown, 1979), and parental separation/divorce. It was hypothesized th at CSA, SES, PBI, and parental separation/divorce would significantly predict BS I scores. It was also hypothesized that CSA would significantly pred ict BSI scores beyond the variance accounted for by the other variables. Results indicat ed that all predictor variable s were significantly related to BSI score in the hypothesized direction, excep t for childhood SES which was found to be unrelated to BSI score in adulthood. A regr ession model including parental care, overprotection, and divorce/separation significan tly predicted BSI score. When objective and subjective CSA severity scores were adde d to the equation, the amount of variance in
v BSI score accounted for signi ficantly increased. Amounts of shared variance were quite high, but results indicated that CSA severity accounts for variance in adult psychological functioning beyond that accounted for by pare ntal care, overprotection, and divorce.
1 Introduction The study of childhood sexual abuse (CSA ) has produced a significant body of research regarding prevalence, risk factors, outcomes, and treatments related to CSA. Throughout the literature, the prevalence of CSA in community samples usually falls between 12% and 35% for women and between 4% and 9% for men (Putnam, 2003). The higher rates are usually found in studies that use a more liberal definition of what constitutes CSA. Significant risk factors for CSA include female gender, older age at the time of abuse, mental and/or physical disabi lity, and parental dysfunction. Outcomes of CSA are usually studied by assessing symptoms of psychological disorders. There are a wide range of symptoms th at have been associated with CSA throughout childhood and adulthood, but the most common across the literature are depression in adulthood and sexualized behaviors in childhood. Empirical study of the long-term outcomes of CSA has shown somewhat inconsistent relationships between experiencing CSA and developing symptoms of psychopathology in adulthood (Neumann, Houskamp, Pollock, Briere, 1996; KendallTackett, Williams, & Finkelhor, 1993; Putn am, 2003). This area of research is complicated by high levels of co-occurre nce between CSA and other childhood risk factors such as physical abuse, emotional abus e, neglect, inter-familial conflict, substance abuse, low socio-economic status, parental psychopathology, family discord, parental separation, and foster care (Maker, Kemme lmeier, & Peterson, 1999; Melchert, 2000; Spaccarelli, 1994). These co-occurring risk fact ors complicate the research process by making both general conceptua lizations and actual statistical analyses more difficult
2 (Romans, Martin, and Mullen, 1997). Co-occurr ing risk factors are difficult to study because a causal chain of events is often impossible to distinguish. Regression analyses are commonly used to quantify risk factors ability to predict ne gative outcomes. When studying CSA and early family environment, high levels of covariation can suppress effect sizes and statistical sign ificance in regression analyses or go so far as to invalidate the analysis. While many of these probl ems cannot be entirely solved, modern experimental design and statis tical analyses do provide the means to examine CSA and other risk factors simultaneous ly, allowing for the comparison of predictive ability of many risk factors for adult psychopathology. The difficulties in studying CSA have resu lted in much of the existing empirical research to be conflicting. Many studies have found relationships of varying strength between childhood sexual abuse and psychopathology symptoms and diagnoses (Kendall-Tackett, Williams, & Finkelhor 1993). Both short-term and long-term outcomes of CSA have been studied, and CSA has been shown to have an effect on the victims psychological health in both chil dhood and adulthood. Research has shown the effects of CSA during childhood to be numerous and varied, unable to be explained with any single symptom or diagnosis (ex. Merry & Andrews, 1994; Koverola, Pound, Hegar, & Lytle, 1993; Oates, OToole, Lync h, Stern, & Cooney, 1994). Merry & Andrews (1994) studied a group of children who were CSA victims, aged 4-16, 12 months following initial disclosure of abuse. They found that these children showed exceptionally high rates of Oppositional Defian t Disorder (ODD), Post-Traumatic Stress Disorder (PTSD), anxiety disorders, depr essive disorders, and Attention-Deficit Hyperactivity Disorder (ADHD), with 63.5% of the children warranting an axis-I
3 diagnosis using the DSM-III-R. In a simila r study Koverola, Pound, Hegar, and Lytle (1993) found that CSA was related to depressi on in children. Using the Child Depression Inventory with a sample of 6-12 year-olds, they found a significant relationship between CSA and symptoms of depression in childhood. These results illustrate the diversity of the potential effects of CSA. Some empirica l research has also examined how childhood symptoms related to CSA change over time Oates, OToole, Lynch, Stern, and Cooney (1994) studied the stability of outcomes relate d to CSA in response to treatment. They found that therapy was not rela ted to outcome. However, they did find that quality of family functioning was related to improvement in self-esteem, depression, and behavior. These results support the need for further study regarding the influence of family environment on outcomes related to CSA. A significant amount of research has also gone into studying the long-term effects of CSA on functioning during adulthood. In a review of 45 studies, Kendall-Tackett, Williams, & Finkelhor (1993) found that many studies showed that childhood sexual abuse was related to anxiety, depression, posttr aumatic stress disorder and other clinical diagnoses. In their review, they found acro ss the 45 studies an average of 28% of survivors of childhood sexual abuse showed anxiety symptomology, 33% showed fear symptomology, 53% showed posttraumatic st ress syndrome symptomology, 28% showed depressive symptomology, 18% showed learni ng difficulties, and 37% showed general behavioral problems. 30% of survivors of childhood sexual abuse s howed symptoms of internalizing disorders while 23% showed symp toms of externalizing disorders, however, these averages were not desc riptive of all 45 studies beca use the range of symptomology was very large. For example, averaged acr oss eight studies, 28% of CSA survivors
4 exhibited anxiety symptomology, however, one study included in this average found only 14% while another found 68% of survivor s of CSA exhibit anxiety symptomology. General PTSD symptoms were reported by an average of 53% of victims of CSA, but the results ranged from 20% to 77%. Suicidal id eation was reported in an average of 12% of CSA survivors with a range of 0% to 45%. Somatic complaints in 14% of survivors ranged from 0% to 60%. Inappropriate sexual behavior was shown in an average of 28% of survivors with a range across studies from 7% to 90%. Self-injurious behavior was shown in an average of 15% of survivors w ith a range between 1% and 71%. Most of these means and ranges were based on the re sults from approximately five studies, indicating that the ranges are not large because of one or two extreme outliers compared to a homogeneous group, but because across st udies there is heterogeneity, a vast difference in results. Some studies linking CS A to the development of symptoms of adult psychopathology show a strong relationship, wh ile others show no relationship (Young, Harford, Kinder, & Savell, 2007). The range in results between studies is extremely problematic to the study of CSAs influe nce on the development of psychopathology. Many studies found very small or no effects of CSA while others found huge effect sizes and a high prevalence of psychopathology symptomology. These wide ranges in results are likely to be partly due to differences in study samples. Great care is necessary when co mparing the likelihood of symptomology in college students, psychiatric inpatients, clin ical outpatients, and community samples due to the differences in the likelihood of ps ychopathology. Another cont ributing factor to this wide range is the definition of CSA. Without a standard definition of CSA, metaanalysis in this area of study becomes questionable. The meta-analysis by Kendall-
5 Tackett, Williams, & Finkelhor (1993), how ever, did show that on average, CSA is related to many symptoms of psychopat hology. This broad range of symptomology, however, indicates that there is no single dia gnosis or type of symptom that can explain the effects of childhood sexual abuse. This ev idence indicates that attempting to define the outcomes of CSA as a specific psychiat ric diagnosis is not empirically supported. Given the difficulty in studying CSAs pred iction of adult psychopathology due to high levels of covariation with other risk factors and the mixed result s of past studies of CSA and adult psychopathology (ex. KendallTackett, Williams, & Finkelhor, 1993; Romans, Martin, and Mullen, 1997; Neuma nn, Houskamp, Pollock, Briere, 1996; Young, Harford, Kinder & Savell, 2007), it is clear that further study is necessary. Ideally, other co-occurring risk factors woul d be included in the analysis. Socioeconomic status (SES) is a vital aspect of early family environment and one su ch co-occurring risk factor. Children and adolescents who come from high SES homes enjoy many advantages and opportunities that low SES homes often do not provide. Because there are many negative outcomes related to low SES, it is one of the most commonly controlled variables in psychological data analysis. In a very inform ative and broad review of the correlates of SES in childhood, Evans (2004) found many relati onships relevant to the study of early family environment predicting adult psychopa thology. While his article did not make a direct link between low childhood SES and adult psychopathology, it does show many specific relationships between SES and other risk factors for nega tive outcomes such as adult psychopathology. If low SES and other risk factors co-occu r with CSA, the simultaneous analysis of these risk factors may be very informativ e and perhaps provide insight into a key limitation to the current body of research on the effects of CSA. Evans
6 showed that as SES increased, the likelihood of the parents being divorced or separated decreased. A child with divorced parents is at increased ri sk due to the lack of both parents in the home, which is related to decreased parental social support, increased interparental conflict, and decreas ed household income due to a parent (and their income) being absent. Evans also found that children in lower SES households were disciplined more harshly and that in early family environm ent, as SES increases, so do mother social support, mother warmth, and cognitive stimulation. Examining a more specific form of cognitive stimulation, Evans (2004) found that children of professional-level parents addressed significantly more words to their children than did working-class parents, who in turn addressed significantly more words to their children than welfare receiving parents. In line with public opi nion, Evans found that SES also influenced the quality of the schools children attended and th e houses they lived in. Overall, Evans study shows us that chil dren of lower SES are more likely to live in households with more conflict, less s upport, less cognitive stimulation, less communication, and lower quality of housing. Also, they are less likely to have adequate facilities for school. Combin ing all these factors sugge sts how many inter-related disadvantages low SES children and adolescent s can face. The co-occurrence of SES risk factors combined with the other early childhood environment risk factors for adult psychopathology suggest that when studying risk factors empirically, it is advantageous to examine many aspects of early family environment simultaneously. SES provides a quantifiable variable that ma y provide information on many co -occurring risk factors that are far more difficult to measure and quantify.
7 Despite substantial evidence of the disa dvantages related to growing up a child with low SES, the current body of literature on CSA and its effects is very limited in its examination of the role that SES may play in the relationship between CSA and negative outcomes in adulthood. Although low socioeconomi c status (SES) is a significant risk factor for physical abuse and neglect, research has shown th at it is a much less powerful predictor of CSA (Putnam, 2003), indicating that CSA is equally, or close to as likely to occur in high or low SES households. Because childhood SES does not appear to predict CSA, the influences of these risk factor s on the development of adult psychopathology are possibly independent. Only by studying thes e variables simultaneously can research hope to show the relationship and possible intera ction between them. To date, this area of research has been understudied, but has shown that childhood SES can play a role in the long-term outcomes related to CSA. In a sample of 90 university clin ic outpatients aged 18-40, among survivors of CSA, high SES was a predictor of better mental health in adulthood (Katerndahl, Burge, & Kellogg, 2005). Along with high SES, lack of family alcohol abuse, fewer abuses by first perpetrato r, and fewer perpetrators predicted better mental health. Porter, Lawson, & Bigler (2005) studied the cognitive abilities and psychopathology of CSA survivors, aged 8 14 at the time of th e study, and found that abuse survivors had higher levels of psychopathology, lower performance on attention/concentration tasks, and lower pe rformance on memory tasks. When controlling for SES and IQ, however, the difference in performance on the memory task became nonsignificant, an example of how negative outcomes associat ed with CSA can sometimes be explained by confounding variables. This re search shows evidence that SES plays a role in the relationship between CSA and negative adult outcomes. High childhood SES
8 appears to protect CSA survivors from negative outcomes while low childhood SES appears to exacerbate the risk. In addition to childhood SES playing a predictive role along with CSA, it appears that CSA may predict adult SES. Romans, Martin, and Mullen (1997) found that women who were victims of CSA we re more likely to have a lower SES than their family of origin. This effect was larger as severity of CSA increased, however, CSA predicting lower SES does not necessarily mean that childhood/adolescent SES and CSA will be related. It is clear that both are risk factors for adult psychopathology and therefore, are wo rth studying together. If CSA and SES both put a person at risk for the development of adult psychopathology, then perhaps the two together will exacerbate the risk, causing th e results of the present study to show an interaction effect. Parental characteristics such as bonding, car e, level of protecti on, parenting style, and inter-familial conflict can also influence the development of adult psychopathology (Chambers, Power, & Durham, 2004; Fosse & Holen, 2006; Heider, Matschinger, Bernert, Alonso, & Angermeyer, 2005; Hill et al., 2000). Like the other risk factors already mentioned, parental ch aracteristics likely share high levels of covariation with other significant risk factors, such as SES. As already di scussed, low SES households are more likely to have divorced or separated parents (Evans, 2004). Divorced or separated parents are more likely to have high levels of inter-parenta l conflict. Also, a child of divorced or separated parents may have less parental support due to the absence of a parent. Empirical research has shown a relationship between divorce and adult psychopathology. Ge, Natsuaki, and Conger (2006) studied the influence of divorce on adolescence and early adulthood depressi on. They found that among both males and
9 females, those in divorced families showed higher levels of depression in late adolescence. This difference remained signi ficant into early adulthood for males, but disappeared for females. Also, depression sc ores were significantly higher for females than males across adolescence and early adulthood. Even when parents are together, there a se veral characteristics of parenting that can lead to an increased likelihood of de veloping adult psychopathology. Enns, Cox, and Clara (2002) used the Parental Bonding Inst rument (PBI; Parker, Tupling, & Brown, 1979) to study the relationship between parental bonding and a dult psychopathology. They found that lack of parental care by bot h mothers and fathers significantly predicted lifetime onset of many forms of psychopat hology, including mood, anxiety, substance use, and personality disorders. Parenting char acteristics as measured by the PBI predicted approximately 1-5% of the variance in adult psychopathology. Because of the apparent link between many aspects of parenting and the developmen t of adult psychopathology, it is important to include parenting variable s in any analyses of childhood and adolescent experiences predicting adu lt psychopathology outcomes. Many studies have examined the rela tionship between parental care and overprotection and their rela tionship to psychological di stress in adulthood. In a comparison of bulimic and non-bulimic partic ipants among psychiatric outpatients, Fosse & Holen (2006) found that those diagnosed w ith bulimia nervosa were more likely to report CSA, emotional abuse, physical abus e, and bullying by peers during childhood. Those diagnosed with bulimia nervosa also scored significan tly higher on father overprotection scale of the PBI, and significan tly lower on the father care scale. In a similar study, Romans, Gendall, Martin, a nd Mullen (2000) found that both CSA and
10 parenting characteristics meas ured by the PBI predicted eat ing disorders in adulthood. Low maternal care specifically predicted anorexia nervosa. Also, among the female survivors of CSA in the sample, paternal ove rprotection and early maturation emerged as significant risk factors for eating disorders in adulthood. When taken in combination, these results support CSAs relationship to negative adult outcomes. The results also support the interaction hypothesi s of the current study; that participants experiencing CSA, low parental care, and high parental overprotection during childhood will be at especially high risk for negative psychological symptoms in adulthood. When studying parenting care and protection with instruments such as the PBI, great care must be taken due to the complexity of the relationship between parenting variable, CSA, and adult outcomes. It appear s that not only do both parental care and CSA predict adult outcomes, but also that parental care can predict CSA. Hill et. al. (2000) found that low maternal and paternal care increased the likelihood of abuse by a non-family member perpetrator before the age of 11, while both maternal care and experiencing CSA predicted adult affective disorder symptoms. The study of CSA has yielded mixed re sults across studies. While some find sexual abuse status alone acc ounted for a very large percen tage of the variance, (ex. 43% for aggression and sexualized behavior s; Kendall-Tackett, Williams, & Finkelhor, 1993), concluding that CSA almost necessarily predicts negative outcomes, others find that CSA is at best a risk f actor among many others and is n either necessary, sufficient, nor acting alone ( Romans, Martin, & Mullen, 1997). A highly likely explanation for the apparent disparity in findings is that CSA is highly interrelated with m any other risk factors for negative outcomes. Empirical rese arch should attempt to disentangle this
11 relationship, simultaneously examining as many risk factors as possible. This will allow for a quantification of more individual aspect s as well as an overall contribution of early family environment in predicting negative outcomes such as adult psychopathology. Some studies have examined the predicti on of negative outcomes in adulthood by both CSA and early family environment. These st udies have produced mixed results. Merrill, Thomsen, Sinclair, Gold, & Milner (2001) found that both CSA and childhood parental support independently and significantly pr edicted adult adjust ment, although this relationship was mediated by coping style. In a similar study, Fassler, Amodeo, Griffin, Clay, & Ellis (2005) found that both severi ty and dichotomous measurements of CSA (abused or not abused) significantly predicted adult outcomes, as did family environment variables. The family environment variab les included conflict, expressiveness, and cohesion and added significantly to the pred ictive power of the regression model beyond the variance accounted for by the CSA variable s. Both studies emphasized the necessity of studying CSA and family environment simu ltaneously in order to maximize clinical utility and our understanding of the factor s contributing to ad ult functioning. Although these studies have found that both CSA and family environment variables can uniquely contribute to the prediction of adult functioning, not all research supports this conclusion. A study by Higgins & McCabe (1994) found that CSA did not significantly contribute to the prediction of adult adjustment beyond th e prediction by family environment. Even though results from previous research ar e in disagreement about CSA and family environments unique prediction of adult outcomes, all their findings support the necessity of simultaneously st udying CSA with other early ex perience variables such as early family environment.
12 Hypotheses 1. It was hypothesized that childhood SES, parental bonding, and parental separation/divorce would significantly pred ict adult psychological functioning individually and as a group. 2. It was hypothesized that the CSA vari ables would predict adult psychological functioning individually and together. 3. It was hypothesized that the CSA va riables would account for a significant amount of variance in adult psychological functioning when added to the prediction model of the parenting variables. 4. It was also hypothesized that a signifi cant interaction woul d be discovered so that children who were sexually abused and in lower SES families would be at highest risk for disorders in adult psychological functioning, while non-sexually abused children with high childhood SES would be at the lowest risk for disordered adult psychological functioning.
13 Method Participants A total of 290 undergraduate females at th e University of South Florida took part in the study. The average age of the sample was 20.4 years ( SD = 2.4) with a minimum age of 18 and a maximum age of 35 years. The sample was 53.3% Caucasian, 19.2% African American, 15.0% Hispanic, 4.2% Asia n American, 4.9% multiracial, and 3.5% other. With regards to roma ntic relationship involvement, 49.1% reported being single, 43.9% were in a romantic relationship, 4.5% were engaged, and 1.0% were married. 34.9% of participants reported that their parents were divor ced. The average participants age at the time of this divorce was 3.1 (SD = 4.7) with a range from before birth to age 19. For taking part in the study, all participants r eceived extra credit to apply to their coursework. There were no limitations on who pa rticipated in the study other than they were female and between the ages of 18 and 35. Measures Demographics were determined by using a demographics questionnaire (appendix B) that asked participants their age, race/ ethnicity, romantic relationship status, and whether or not there was any parental divor ce/separation before the age of 18. For the purposes of assessing childhood socioeconomic status, the demographic questionnaire also asked the participants primary childhood and adolescence caretak ers (parents or guardians) occupation and level of education.
14 Current psychopathology symptoms were assessed with the Brief Symptom Inventory (BSI; Derogatis, 1982), a 53-item self-report measure designed to assess common symptoms of psychopathology. Respondent s were asked to rate the extent to which each item/problem has distressed them over the past seven days. Answers are on a 5-point Likert scale ranging fr om not at all to extremely. The BSI consists of nine subscales, which include depression, interpersonal sensitivity, anxiety, phobic anxiety, paranoid ideation, somatization, obsessive-compu lsive, hostility, and psychoticism. The BSI has demonstrated good reliability, with in ternal consistency values for the subscales ranging from .71 for the psychoticism subscale to .83 for the obsessive-compulsive subscale. The BSI has also been reported as ha ving test-retest reliability values of above .80 for the global severity index (Mental Measurements Yearbook, 1990). Childhood sexual abuse was assessed using the Early Sexual Experiences Survey (ESE; Bartoi & Kinder, 1998). The ESE was mo dified for the purposes of this study, to add a subjective classification question (descr ibed below), and can be found in Appendix A. This measure defines CSA as any sexual contact between a child under the age of 16 and someone at least five years older. The ESE is a 14-item measure that asks respondents to indicate whether or not th ey experienced various types of sexual encounters before the age of 16 using a yes (1) or no (0) format. A participant responding no to all of the first ten items will be treated as having no history of CSA while a participant who responds yes to any of the first ten items on this scale will be treated as meeting objective criteria for a hist ory of CSA. For participants with a history of CSA, the total number of yes responses will be used to produce an objective CSA severity score ranging from 1-10, with 1 bei ng the least severe a nd 10 being the most
15 severe. A subjective CSA severity classifica tion and score was created with items 11 and 12. Item 11 asks participants Do you consider yourself to be a victim of childhood sexual abuse? providing a subjective classi fication as abused or non-abused. Item 12 asks participants to rate how severely the experience (any yes to items 1-11) impacted their life (0 being no negative impact at all to 10 being a severe negative impact), providing a subjective severity measure. The ESE has adequate reliability with reported internal consistency values around .79 (Young, Harford, Kinder, & Savell, 2007). Childhood socio-economic status (SES) wa s computed using the Hollingshead (1975) system, which approximates childhood SES with parental education levels and an occupation score. Education is rated from 1 to 7 with 1 equal to less than a seventh grade education through 7 equal to graduate training. Occupations are scored from 1 to 9 with 1 equal to occupations such as farm laborers or menial service work ers through 9 equal to occupations such as executives, proprietors of large businesses, or major professionals. Education and occupation scores are then weighted and combined into a total score, ranging from 8-66. For families with multiple incomes/caretakers, the total scores are averaged to get a single SES score for the family. Parental bonding was asse ssed using the Parental Bonding Instrument (PBI) developed by Parker, Tupling and Brown (1979; appendix C). This 48-item questionnaire assesses two aspects of parental bonding, car e and overprotection, by asking participants retrospectively about their ch ildhood experiences with their parents. These factors were defined by factor analysis. In a 20-year longitudinal study, Wilhelm, Niven, Parker, & Hadzi-Pavlovic (2005) reported the PBI has adequate psychometric properties. They found the maternal care subscal e of the PBI has a test-retest reliability of .75 over a 5-
16 year period, .64 over a 10-year period, and .73 over a 20-year period. They reported the maternal overprotection subscale as having a test-retest reliability of .75 over a 5-year period, .67 over a 10-year period, and .69 over a 20-year period. They reported the paternal care subscale as ha ving a test-retest re liability of .82 over a 5-year period, .74 over a 10-year period, and .75 over a 20-year period. They reported the paternal overprotection subscale as having a test-retest reliability of .74 over a 5-year period, .62 over a 10-year period, and .59 over a 20-year period. The PBI was used to assess the extent to which, during a participants childhood/adolescence, parents were overprotecting vs. allowing of autonomy and caring vs. indifferent/rejecting. Procedure Participants first completed the informed consent form followed by the demographics questionnaire, the Early Sexua l Experiences Questionnaire, the Parental Bonding Instrument, and the Brief Symptom I nventory in a random order. All measures were filled out in one session. Participants filled out the questionnaires in groups with spacing adequate to ensure individual pr ivacy of responses. Informed consent and questionnaire packets were kept separate from each other, and the informed consent forms were shuffled upon receipt to ensure th at an informed consent form could not be matched with its corresponding questionnaire packet. Upon comp letion, participants were thanked, debriefed, given a chance to ask quest ions and express any concerns, and given referral sources if any adverse effects were experienced.
17 Results The mean BSI score for the entire sample was 43.11 ( SD = 34.57) with a maximum of 187. The PBI subscale score means for the entire sample were as follows; father care was 24.08 ( SD = 10.11), father overprotection was 15.36 ( SD = 7.39), mother care was 29.93 ( SD = 7.30), and mother overprotection was 14.89 ( SD = 7.71). Participants were identified as having experienced CSA by the ESE-R. Of the 290 participants, 39.3% reported experiencing at l east 1 incident of CSA before the age of 16 (endorsing yes on at least 1 ESE-R item 1-10) and were classified as experiencing CSA using the objective classification. Of these participants, 50.9% re ported having sexual contact with someone at leas t 5 years older than them; 32.5% reported being forced into sexual activity by a perpetrator of any age; 89.5% reported be ing touched in a way that made them feel violated; and 14.9% reported engaging in unwanted sexual activity while too intoxicated or influenced by drugs to give consent. However, when asked do you consider yourself to be a victim of CSA, only 10.0% of the 290 participants answered yes. Only 9.6% of the objectively identifie d abused participants reported receiving psychological treatment in which sexual a buse was one of the issues covered. Of particular note, these results show that through objective iden tification, 39.3% of participants were identified as experienci ng CSA while through subjective identification, only 10% were. Using an objective and subjectiv e severity score is advantageous because it addresses the problem of the large difference between objective and subjective identification rates. The follo wing analyses were conducted with data from the entire
18 sample. The sample size was sufficient that participants with missing data were simply removed from the analysis. This resulted in some variation of sample sizes. It was hypothesized that the CSA variab les would significantl y correlate with each other and with the BSI tota l and subscale scores. To test this hypothesis, Pearsons correlation coefficients were computed between the CSA variables, the BSI total score, and the nine BSI subscale scores. The resu lts supported the hypot heses. The objective CSA severity score and subjective CSA seve rity rating were signi ficantly correlated ( r = .650, p < .001). The objective severity score was al so significantly co rrelated with BSI total score ( r = .255, p < .001) and all 9 BSI subscales (minimum r = .144, p = .015 for the interpersonal sensitivity subscale, maximum r = .255, p < .001 for the somatization subscale). The subjective sever ity rating was also significantly correlated with BSI total score ( r = .251, p < .001) and all 9 BSI subscales (minimum r = .163, p = .005 for the interpersonal sensitivity subscale, maximum r = .262, p < .001 for the anxiety subscale). These results are summarized in table 1.
19 Table 1. Correlation matrix of CSA variables and BSI subscales 1 2 3 4 5 6 7 8 9 10 11 12 1.CSA objective severity -.65** .26* .15* .14* .24** .23** .16* .18* .23** .23** .26** 2.CSA subjective severity -.24** .17* .16* .19* .26** .17* .23** .24** .23** .25** 3.Somatization -.68** .54** .63** .77** .59** .59** .55** .64** .82** 4.Obs. Compulsive -.68** .69** .68** .59** .55** .63** .67** .84** 5.Interpersonal Sensitivity -.74** .63** .62** .56** .72** .73** .83** 6. Depression -.71** .64** .54** .69** .82** .88** 7.Anxiety -.67** .65** .62** .73** .87** 8.Hostility -.45** .60** .62** .76** 9.Phobic Anxiety -.53** .58** .70** 10.Paranoid Ideation -.72** .81** 11.Psychoticism -.87** 12.BSI Total -* p < .01, ** p < .001 Before conducting analyses using the Hollingshead childhood SES score, a oneway ANOVA was conducted to test for race/ ethnicity group differences on the SES score. This test revealed that there were significant SES differences between some races/ethnicities ( F (5, 275) = 2.393, p = .038), therefore, race/ethni city was held constant for any analyses including SES. It was hypothesized that significant correlations would be found between the Hollingshead childhood SES score, the PBI subscal e scores, and BSI total score, so that higher SES is related to better (higher wa rmth, lower overprotection) PBI scores and lower BSI scores, and that better PBI scor es are related to lower BSI scores. This hypothesis was tested by computing a Pears ons correlation matrix including childhood
20 SES, PBI subscale scores, and BSI total sc ore. The results indicate that, of these variables, childhood SES was only significantly correlated with the father care subscale of the PBI ( r = .194, p = .003). BSI score was significan tly correlated with all 4 PBI variables in the hypothesized direction, but not with chil dhood SES (table 2). Because SES was not even marginally correlated with BSI total score ( r = -.005, p = .993), it was dropped from all further analyses. It was also hypothesized that parental divorce/separation during childhood would be related to higher BSI score. To test this hypothesis, a one-way ANOVA was conducted. The results indicate that parental divorce/separation during childhood was only marginally related to BSI score ( F (2,280) = 2.768, p = .064). The results of these analyses in dicated that objective CSA severity, subjective CSA severity, the 4 PBI subscales and parental separation/divorce are all related to psychological functioning in a dulthood. These variables were therefore included in the regression analysis. Table 2. Correlation matrix of SES, PBI subscales, and BSI total controlling for race/ethnicity 1 2 3 4 5 6 1. SES -.194** -.028 .087 .005 -.005 2. Father Care --.322*** .277*** -.032 -.128* 3. Father O.P. --.241*** .399*** .181** 4. Mother Care --.257*** -.157** 5. Mother O.P. -.283*** 6. BSI Total -* p < .05, ** p < .01, *** p < .001 It was hypothesized that the predictor variables, objective CSA severity, subjective CSA severity, childhood SES, th e PBI subscale scores, and parental separation/divorce, would significantly predic t BSI total score individually and as a
21 group, and that the CSA variables would si gnificantly predict BSI score beyond the prediction made by the other variables. SES was not included in this analysis once it was found to be unrelated to BSI score. Multiple regression analyses were conducted to test these hypotheses. As hypothesi zed, the model including the 4 PBI subscales and parental separation/divorce significantly predicted BSI score ( R = .106, F (5,237) = 5.601, p < .001; table 3). The model including only the CSA objective and subjectiv e severity scores also significantly predicted BSI score ( R = .078, F (2,278) = 11.801, p < .001; table 4). Also as hypothesized, the model including the parental variables, objective CSA severity, and subjective CSA severity sign ificantly predicted BSI score ( R = .174, F (7,233) = 6.994, p < .001; table 5). An R change test between the 2 regression models (F(2,232) = 9.550, p < .01) indicated that the CSA variables do add to the predic tive ability of the parenting variables. However, when examin ing the individual contributions of the CSA variables to the model, only the subjectiv e severity score accounts for a significant amount of unique variance beyond that accounted for by the other variables, ( = .186, p = .021). Table 3. Regression of BSI total score on PBI s ubscales and parental divorce/separation b p Total Model Father Care -.243 -.068 .332 R = .106, F(2,237) = 5.601(5,237), p < .001 Father O.P. .161 .033 .643 Mother Care -.344 -.072 .275 Mother O.P. 1.111 .239 .001 Parental Divorce 6.796 .106 .099
22 Table 4. Regression of BSI total score on CSA objec tive severity and CSA subjective severity b p Total Model CSA Obj. 3.297 .158 .038 R = .078, F (2,278) = 11.801, p < .001 CSA Subj. 2.217 .151 .048 Table 5. Regression of BSI total score on PBI subs cales, parental di vorce/separation, CSA objective severity, and CS A subjective severity b p Total Model Father Care -.161 -.045 .514 R = .174, F (7,233) = 6.994, p < .001 Father O.P. .133 .027 .693 Mother Care -.225 -.047 .466 Mother O.P. 1.003 .215 .002 Parental Divorce 5.646 .088 .158 CSA Obj. 2.391 .109 .173 CSA Subj. 2.885 .186 .021
23 Discussion This study attempted to add to the exis ting body of CSA literature by examining CSAs relationship with adult psychologica l functioning while simultaneously analyzing co-occurring childhood factors such as SES, parental care, parental overprotection, and parental divorce. It was hypot hesized that lower SES, lo wer parental care, greater parental overprotection, and gr eater rate of divorce woul d all be associated with disordered adult psychological functioni ng. A major limitation of the CSA body of research is the lack of examination of co-occurr ing risk factors. This has led to debate as to whether CSA leads to psychopathology in adulthood, or whethe r this relationship would be better accounted for by other chil dhood risk factors. Many studies have disagreed on the exact nature of this relationship, but they have reached consensus on the necessity of examining co-occurring risk and protective factor s (Merrill, Thomsen, Sinclair, Gold, & Milner, 2001; Fassler, Amodeo, Griffin, Clay, & Ellis, 2005; Higgins & McCabe, 1994). The current study examined CSA in 2 ways, both different than the majority of existing studies. Commonly, participants are placed in abus ed or non-abused groups based on their endorsement of questionnaire items, or answers to interview questions. Most commonly, if a particip ant indicates having experi enced any form of childhood abuse, they are classified as abused. This approach has 2 main problems. The first is that what constitutes abuse varies greatly across studies. As discussed previously, this may be a reason for the range of results from one study to another. If only very severe abusive events are used to classify participan ts as abused, then it is more likely that the
24 prevalence of CSA will be lower and the out comes will likely be more severe. The second problem with this approach is comm on to any study that condenses scores into dichotomous variables, and is th at a great amount of variance is lost in the process. Using severity scores rather than dichotomiz ing abused and non-abused provides the advantages of more descriptive data and greater statistical power. The majority of existing studies also make abuse classifi cations based only on objective information, what abusive events actually occurred. This approach neglects the subjective nature of the victims own thoughts and feelings surround ing the abuse. For example, if a 15-yearold girl has a consensual sexual relationship with a 21-year-old ma n, many studies would classify this as abuse (indeed this woul d count towards the cu rrent studys objective severity score). If the girl later felt taken a dvantage of, she may subjectively feel that she was abused and therefore experience gu ilt, depression, or any other negative psychological outcome that has been found to be related to CSA. However, if she goes on with her life, always thinking of the relationship as ha ving been healthy and consensual, she may never suffer a negative outcome. Usi ng a subjective severity rating of abuse provides information that may be missed wh en using only objective classification. The limitation of using only subjective severity ratings is that participants may be more likely to assign greater severity to past events if they are currently experiencing depression, anxiety, etc. This may artificially inflat e the relationship between subjective CSA severity and current symptomology. In an at tempt to maximize accuracy, descriptiveness, and statistical power while minimizing disadva ntages, the current study measured CSA in 2 ways; an objective severity score and a subjec tive severity score. The objective score is simply how many abusive events the subject reported experiencing, while the subjective
25 score is a rating on how much any abuse ne gatively impacted their lives (ESE-R; appendix A). Results indicated th at, although highly correlated ( r = .650, p < .001), the objective and subjective severity ratings do appear to be m easuring different constructs. Analysis of the objective and subjecti ve CSA severity scores confirmed the hypothesis that they would be related to adult psychological functioning. While most prior research on CSA (ex. Kendall-Tackett, Williams, & Finkelhor, 1993) has shown that those having experienced CSA are mo re likely to have psychopathology in adulthood, the results of the current study ex tend these findings by showing that both objective and subjective severi ty ratings are positively co rrelated with number of symptoms of psychopathology. As severity of abuse, both objectiv ely and subjectively measured, increases, adult psychological func tioning becomes more impaired. This result was found for all the BSI subscales; depressi on, interpersonal sensitivity, anxiety, phobic anxiety, paranoid ideation, somatization, obsessive-compulsive, hostility, and psychoticism. Results of the current study showed th at the parental care and overprotection variables were all sign ificantly correlated with adult psychologi cal functioning. This finding supports the hypotheses and agrees with past research (ex. Enns, Cox, and Clara, 2002). It appears that higher parental care dur ing childhood acts as a protective factor for developing psychopathology in ad ulthood. It also appears that parental overprotection puts children at higher risk for developi ng psychopathology in adulthood. Parental divorce before the child turned 18 was found to be marginally related to BSI score, possibly increasing the risk of psychopat hology during adulthood. This finding agreed with previous findings by Ge, Natsuaki, and Conger (2006), th at parental divorce
26 predicted depressive symptoms among adolesce nts, but that the effect disappeared in adulthood for women. Since the current study used a sample of young women, this marginal finding may be explained by the young age of participants, many of them having recently b een adolescents. SES has been linked with many risk a nd protective factors in childhood. Although Evans (2004) found that higher SES was associated with many childhood advantages including greater parental s upport and greater cognitive stimu lation, he did not discuss a link between lower childhood SES and ps ychological functioning. It had been hypothesized that childhood SES would be related to, and perhaps even interact with, CSA to predict adult psychological functioning. As SES was not even marginally related to adult psychological functioning, the hypot hesis was not supported and SES was dropped from all analyses. One potential e xplanation for this finding is that the participants were all undergr aduate students at a major uni versity in Florida. Simply being college students limits the range of childhood SES because having low SES limits college attendance. This range restriction could account for the findings. Another possibility is that childhood SES is not re lated to adult psychological functioning as measured by the BSI. The BSI assessment tool focuses on symptoms and traits related to psychopathology. It appears that childhood SES is not related to adult psychopathology among people who are of high enough SES to go to college. It is possible that a replication of this study using a more economically diverse sample would find a relationship between childhood SES and adult psychological functioning. It is also possible that a study measuring another psyc hological outcome variable, such as IQ score, would find a relationship with childhood SES. For the current study, any
27 hypothesis regarding SES was not supported and analysis was limited to the other variables. This study attempted to contribute to the body of CSA literature by assessing multiple childhood factors simultaneously, in cluding parenting variables and CSA, allowing the researchers to quantify and compar e the predictive power of these factors on adult psychological functioning. Th e results of the correlationa l analysis established that all of the predictor variables could significantly predict BSI score. The larger regression models were constructed to show the predictive power of these variables as a group, and examine any overlap in predicted variance. The results showed that a regression model including mother care and overprotection, fa ther care and overprot ection, and parental divorce, accounted for approximately 11% of va riance in adult psychological functioning as measured by the BSI. The amount of overl ap was very high among these variables, with only mother overprotection accounting for a significant amount of unique variance. The model including only the abuse severity variables accounted for approximately 8% of the variance in adult psyc hological functioning. Both object ive and subjective severity scores accounted for a significant amount of unique variance in this model. The model including all predictor variables accounted for approximately 17% of the variance in adult psychological functioning. Again, th e amounts of overlap were very high. Subjective, not objective, CSA severity sc ore accounted for a significant amount of unique variance in to model with all predicto rs. As hypothesized, the significant increase in the R2 indicates that CSA accounts for variance in adult psychological functioning beyond that accounted for by the parenting variables. Comparing the objective and subjective CSA severity scores provides inte resting information. As already discussed,
28 the correlation between the 2 was moderately high, but not high enough to indicate that the variables measured exactly the same cons truct. Results of the regression analysis indicated that the 2 scores did not overlap as much as might be expected (both accounted for a significant amount of unique variance in the regression model using only the CSA severity scores) in the prediction of adult ps ychological functioning. Th erefore, this study supports the use of not only severity ratings (instead of assigning participants to dichotomous groups), but also the use of bot h objective and subjective severity measures of CSA. When examined as a whole, such as w ith prior meta-analysis of 45 studies by Kendall-Tackett, Williams, & Finkelhor ( 1993), the body of literature on CSA does indicate that CSA is related to negative psychological outcomes in adulthood, but these results vary greatly in the st rength of this relationship. Researchers have hypothesized that this was because of diffe ring definitions of CSA, and varying levels of examination of other risk factors ( Fassler, Amodeo, Griffin, Clay, & Ellis, 2005; Higgins & McCabe, 1994; Merrill, Thom sen, Sinclair, Gold, & Milner, 2001; Romans, Martin, & Mullen, 1997; Young, Harford, Kinder, & Savell, 2007 ). The current study addressed these limitations in 2 ways, by using severity scores to address problems with defining abused vs. non-abused, and by examining the prediction of multiple risk /protective factors simultaneously. Addressing the larger issue; does CSA alone necessarily lead to negative psychological outcomes? Kendall-Tackett, W illiams, and Finkelhor (1993) found that among 45 studies, CSA accounted for 15-45% of variance in adult psychological functioning. Given the variation in outcomes of CSA, they conclude, ...the absence of
29 any specific syndrome in children who ha ve been sexually abused, and no single traumatizing process. Their findings seem to indicate that CSA accounts for large amounts of variance in adu lt psychological functioning, but that CSA does not act through a simple process. Rather, it can have influence through many processes and its outcomes vary greatly. In another review, Pu tnam (2003) found that CSA appears to be related to a wide variety of negative ps ychological outcomes in adulthood. Romans, Martin, and Mullen (1997) even go so far as to conclude that CSA is best conceptualized as a non-specific risk factor for a wide range of psychologicaloutcomes. The current study attempted to measure some of the ot her potential co-occurri ng risk factors and quantify their influence in comparison with CSA. Results indicated the parenting variables such as care, overprotection, and divorce are also related to negative psychological outcomes in adulthood, and acc ount for similar amounts of variance (to CSA) in these outcomes. The current findings also indicate that, as hypothesized, the amounts of shared variance be tween the parenting variable s and CSA were quite high, but that CSA did account for a significant am ount of variance in adult psychological functioning beyond that of the other variables. The current study supports using severity sc ores instead of dichotomous groups in future research. It also highlights th e need to examine multiple risk-factors simultaneously. The current study used only fe male participants, therefore, future research should be conducted to examine if these findings generalize to men. Future research should also examine other potenti al childhood factors that may be related to CSA and predict adult psychological functioning. Such factors incl ude education, living situation, parental psychopathology, and many ot hers that could all show results similar
30 to those in this study. Other types of abuse or neglect may also predict psychopathology in adulthood while being related to CSA and other risk factors. Once a great number of risk factors are identified, interventions for victims of CSA will be able to take into account those factors that appe ar to have the most significant impact on the victims.
31 References Bartoi, M. & Kinder, B. (1998). Effects of child and adult sexual abuse on adult sexuality. Journal of Sex and Marital Therapy 24, 75-90. Chambers, J., Power, K., & Durham, R. (2004). Parental styles and long-term outcome following treatment for anxiety disorders. Clinical Psychology and Psychotherapy, 11, 187-198. Cohen, J. (1992). A power primer. Psychological Bulletin 112, 155-159. Derogatis, L. (1982). Brief Symptom I nventory: Administra tion, Scoring, and Procedures Manual-II.. Minneapolis : NCS Pearson Assessments, Inc. Enns, M. W., Cox, B. J., Clara, I. (2002) Parental bonding and adult psychopathology: Results from the U.S. national comorbidity study. Psychiatric Medicine, 32, 9971008. Evans, G. W. (2004). The environment of childhood poverty. American Psychologist,59 (2), 77-92. Fassler, I., Amodeo, M., Griffin, M., Clay, C., & Ellis, M. (2 005). Predicting long-term outcomes for women sexually abused in ch ildhood: Contribution of abuse severity versus family environment. Child Abuse & Neglect 29, 269-284. Fosse, G. & Holen, A. (2006). Childhood maltreatment in adult female psychiatric outpatients with eating disorders. Eating Behaviors, 7, 404-409. Ge, X., Natsuaki, M., & Conger, R. (2006). Trajectories of depressive symptoms and stressful life events among male and female adolescents in divorced and nondivorced families. Development and Psychopathology, 18, 253-273.
32 Heider, D., Matschinger, H., Bernert, S., Alonso, J., & Angermeyer, M. (2005). Relationship between parental bonding and mood disorder in six European countries. Psychiatry Research, 143, 89-98. Higgins, D., & McCabe, M. (1994). The relatio nship of child sexual abuse and family violence to adult adjustment: Toward an integrated risk-sequelae model. The Journal of Sex Research, 31 (4), 255-266. Hill, J., Davis, R., Byatt, M., Burnside, E., Rollinson L., & Fear, S. (2000). Childhood sexual abuse and affective symptoms in women: A general population study. Psychological Medicine, 30, 1283-1291. Hollingshead, A. B. (1975). Four factor index of social status Unpublished manuscript, Yale University, NewHaven, CT. Katerndahl, D., Burge, S., & Kellogg, N. (2005) Predictors of development of adult psychopathology in female victims of childhood sexual abuse. The Journal of Nervous and Mental Disease,\193(4), 258-264. Kendall-Tackett, K., Williams, L., & Finkelhor, D. (1993). Impact of sexual abuse on children: A review and synthesis of recent empirical studies. Psychological Bulletin, 113(1), 164-180. Koverola, C., Pound, J., Heger, A., & Lytle, C. (1993). Relationship of child abuse to depression. Child Abuse & Neglect 17, 393-400. Maker, A., Kemmelmeier, M., & Peterson, C. (1999). Parental sociopathy as a predictor of childhood sexual abuse. Journal of Family Violence, 14, 1999. Melchert, T. (2000). Clarifying the effects of parental substance abuse, child sexual abuse, and parental caregiving on adult adjustment. Professional
33 Psychology:Research and Practice, 31, 64-69. Mental Measurements Yearbook 10th Edition (1990). Lincoln, Nebraska: Buros Institute of Mental Measurements, University of Nebraska Press. Merrill, L., Thomsen, J., Sinclair, B., Gol d, R., & Milner, S. (2001). Predicting the impact of childsexual abuse on women: The role of abuse severity, parental support, and coping strategies. Journal of Consulting and Clinical Psychology 69(6), 992. Merry, S. & Andrews, L. (1994). Psychiatri c status of sexually abused children 12 months after disclosure of abuse. Journal of American Academy of Child & Adolescent Psychiatry 33, 939-944. Neumann, D., Houskamp, B., Pollock, V., Briere J. (1996). The long-term sequelae of childhood sexual abuse in women: A meta-analytic review. Child Maltreatment,1, 6-16. Oates, K., OToole, B., Lynch, D., Stern, A., & Cooney, G. (1994). Stability and change in outcomes for sexually abused children. Journal of American Academy of Child & Adolescent Psychiatry, 33, 945-953. Parker, G., Tupling, H., & Brown, L. ( 1979). A parental bonding instrument. British Journal of Medical Psychology, 52, 1-10. Porter, C., Lawson, J., & Bigler, E. (2005). Neurobehavioral sequelae of child sexual abuse. Child Neuropsychology, 11, 203-220. Putnam F. (2003). Ten year research update review: child sexual abuse. Journal of the American Academy of Child & Adolescent Psychiatry 42, 269-278. Romans, S., Gendall, K., Martin, J., & Mullen, P (2001). Child sexual abuse and later
34 disordered eating: A New Zealand epidemiological study. Journal of Eating Disorders, 2 9, 380-392. Romans, S., Martin, J., & Mullen, P. ( 1997). Childhood sexual abuse and later psychological problems: Neither necessa ry, sufficient, nor acting alone. Criminal Behavior and Mental Health, 7(4) 327-338. Spaccarelli, S. (1994). Stress, appraisal, and coping in child sexual abuse: A theoretical and empirical review. Psychological Bulletin, 116, 340-362. Wilhelm, K., Niven, H., Parker, G., & Hadzi-Pa vlovic, D. (2005). The stability of the Parental Bonding Instrument over a 20-year period. Psychological Medicine, 35, 387-393. Young, M. S., Harford, K., Kinder, B. N., & Savell, J. K. (2007). The relationship between childhood sexual abuse and adult mental health among undergraduates: Victim gender doesn't matter. Journal of Interper sonal Violence, 22 1315-1331.
36 Appendix A Early Sexual Experiences Surv ey (Bartoi & Kinder, 1998) We would like to get an idea about the type of sexual experiences you may have had before the age of 16 (15 and younger). Please answer yes or no to the following questions in terms of that time. Before the age of 16 (15 and younger) No Yes 1. Did you ever touch the genitals of someone at least 5 years older than you? 0 1 2. Did someone at least 5 years older than you ever touch your genitals or breasts (besides for a physical examination)? 0 1 3. Did you engage in oral sex (cunnilingus and/or fellatio) with someone at least 5 years older than you? 0 1 4. Did you engage in vaginal intercourse with someone at least 5 years older than you? 0 1 5. Did you engage in anal intercourse with someone at least 5 years older than you? 0 1 6. Were you forced into genital manipulation that was unwanted by anyone of any age? 0 1 7. Were you forced into oral sex (cunnilingus and/or fellatio) that was unwanted by anyone of any age? 0 1 8. Were you forced into anal intercourse that was unwanted by anyone of any age? 0 1 9. Were you ever touched in a way that made you feel violated? 0 1 10. Did you engage in any unwanted sexual activity while too intoxicated or influenced by drugs to give consent? 0 1 11. Do you consider yourself to be a victim of childhood sexual abuse? 0 1 12. If you answered yes to ANY of the above questions, please rate the extent to which your experience had a negative impact on your life (0 being no negative impact at all, 5 being a moderate negative impact, and 10 being a severe negative impact; CIRCLE ONE) 0 1 2 3 4 5 6 7 8 9 10 13. Did you ever receive psychological treatment? 0 1 14. If yes, was sexual abuse one of the issues covered? 0 1
37 Appendix B Demographic Information 1. Age in years: _____ 2. Preferred ethnic / r acial designation: African-American (Black) Asian-American Caucasian (White) Latino (Hispanic) Multiracial Native American (Indian) Specify if not listed: ________________________________ 3. Current romantic relationship status: Single Married In a relationship Divorced Engaged 4. Check all the experiences you had before the age of 16: Hospitalization for physical illness Hospitalization for psychiatric illness Major accident or injury Handicap or disability Out-of-home placement Death of parent Parental separation or divorce If you checked box above, please indicate your age when the divorce or separation occurred: ______ Imprisonment of a parent Death of a sibling Loss of a sibling thro ugh separation or divorce Department of Social Services involvement Juvenile justice system involvement Other agency involvement (please specify ________________________) 5. Which of the following best describes your most typical living situation during each of the following age ranges: Birth to 6 Years 7-12 Years 13 Years & Older
38 With both natural parents With a natural parent & a step-parent With a single natural parent With an adoptive parent With a foster family With grandparents or other relatives 6. Number of younger siblings living in the home during each of the following age ranges: Birth to 6 years 7-12 Years 13 Years & Older ____ ____ ____ 7. Number of older siblings living in the home during each of the following age ranges: Birth to 6 years 7-12 Years 13 Years & Older ____ ____ ____ 8. Check all special academic placements you had while in school: None Advanced Placement Gifted and Talented Educationally handicapped Learning disabled Homebound Vocational rehab Other (please specify ________________________) 9. While growing up, did you regularly attend a place of worship? Yes No 10. While growing up, what was the highes t education achieved by your primary parent(s)/guardian(s)? (examples: high school diplom a, GED, bachelors degree, masters degree, PhD) Parent/guardian 1:____________________________________________ Parent/guardian 2:____________________________________________ 11. While growing up, what was your primar y parent(s)/guardian(s) occupation?
39 (if more than 1 occupation, please write in the occupation done for the largest amount of time during your childhood) Parent/guardian 1:____________________________________________ Parent/guardian 2:____________________________________________
Appendix C 40