<%BANNER%>

Family Experience, Attachment to Parents, and Low Self-Control in Dating Violence in the United States and Korea

Permanent Link: http://ufdc.ufl.edu/UFE0024382/00001

Material Information

Title: Family Experience, Attachment to Parents, and Low Self-Control in Dating Violence in the United States and Korea
Physical Description: 1 online resource (183 p.)
Language: english
Creator: Park, Mi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: abuse, american, child, college, comparison, control, country, date, dating, domestic, factor, intimate, korean, learning, partner, protective, race, relationship, risky, self, sexual, students, theory, violence
Criminology, Law and Society -- Dissertations, Academic -- UF
Genre: Criminology, Law, and Society thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: FAMILY EXPERIENCE, ATTACHMENT TO PARENTS, AND LOW SELF-CONTROL IN DATING VIOLENCE IN THE U.S. AND KOREA MiRang Park 352-262-0949 Department of Sociology, Criminology and Law & Society Supervisory chair: Ronald Akers Co-char: Angela Gover Ph.D May, 2009 Although many studies have been done about dating violence, only a few researchers have developed theoretical conceptualizations of factors such as these in dating violence (Sellers, 2005; Follingstad et al, 1999; Riggs and O Leary, 1989; Luthra and Gidycz, 2006). This dissertation moves beyond these to draw on three prominent general theories of crime and deviance for its theoretical framework to understand the effects of early family experience with violence (social learning), current attachment to parents (social bonding), and self-control. The purpose of this study is to link theoretically relevant variables that are linked in some way to the respondents family of origin that may account for the likelihood of perpetration of or victimization by dating violence in two different cultures. In using the same survey instrument for two countries, this study tested whether theoretical explanations developed in Western society could be applied to Eastern society. A sample of 1,400 young adults attending a university in Korea and a sample of 1,500 young adults attending a university in America were used in this dissertation. Findings indicate that experience of victim of child abuse (social learning), low self-control (self-control), serious dating relationship and risky sexual behavior have significant effects on both the perpetration and victimization of dating violence in both countries. Self-control has stronger effects in the American, and being the victim of violence in the home has more significant effects in the Korean sample.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mi Park.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Akers, Ronald L.
Local: Co-adviser: Gover, Angela R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-11-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024382:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024382/00001

Material Information

Title: Family Experience, Attachment to Parents, and Low Self-Control in Dating Violence in the United States and Korea
Physical Description: 1 online resource (183 p.)
Language: english
Creator: Park, Mi
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: abuse, american, child, college, comparison, control, country, date, dating, domestic, factor, intimate, korean, learning, partner, protective, race, relationship, risky, self, sexual, students, theory, violence
Criminology, Law and Society -- Dissertations, Academic -- UF
Genre: Criminology, Law, and Society thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: FAMILY EXPERIENCE, ATTACHMENT TO PARENTS, AND LOW SELF-CONTROL IN DATING VIOLENCE IN THE U.S. AND KOREA MiRang Park 352-262-0949 Department of Sociology, Criminology and Law & Society Supervisory chair: Ronald Akers Co-char: Angela Gover Ph.D May, 2009 Although many studies have been done about dating violence, only a few researchers have developed theoretical conceptualizations of factors such as these in dating violence (Sellers, 2005; Follingstad et al, 1999; Riggs and O Leary, 1989; Luthra and Gidycz, 2006). This dissertation moves beyond these to draw on three prominent general theories of crime and deviance for its theoretical framework to understand the effects of early family experience with violence (social learning), current attachment to parents (social bonding), and self-control. The purpose of this study is to link theoretically relevant variables that are linked in some way to the respondents family of origin that may account for the likelihood of perpetration of or victimization by dating violence in two different cultures. In using the same survey instrument for two countries, this study tested whether theoretical explanations developed in Western society could be applied to Eastern society. A sample of 1,400 young adults attending a university in Korea and a sample of 1,500 young adults attending a university in America were used in this dissertation. Findings indicate that experience of victim of child abuse (social learning), low self-control (self-control), serious dating relationship and risky sexual behavior have significant effects on both the perpetration and victimization of dating violence in both countries. Self-control has stronger effects in the American, and being the victim of violence in the home has more significant effects in the Korean sample.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mi Park.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Akers, Ronald L.
Local: Co-adviser: Gover, Angela R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-11-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024382:00001


This item has the following downloads:


Full Text

PAGE 1

1 FAMILY EXPERIENCE, ATTACHMENT TO PARENT S, AND LOW SELF CONTROL IN DATING VIOLENCE IN THE U.S. AND KOREA By MI RANG PARK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Mi Rang Park

PAGE 3

3 To my family, for sacrificing to provide me with opportuni ties to think, change, and grow

PAGE 4

4 ACKNOWLEDGMENTS I thank my parents and all of my family for believing not so much in the venture, but in the person. I am grateful to my Korean advisors (Dr. Minsik Lee) who mentally supported me to study Criminology in U.S. A. I thank all of all my friends, colleagues and contemporaries in graduate school (especially Seokmin, HeeJung, Suchat, Jeff, Saskia, Krissy,and Q ) for making my work more enjoyable and rewarding than I could have ever expected. I thank the current and former faculty in Criminology, Law and Society at the University of Florida for nur turing a hometown prospect and for always k eeping their office doors open. I am very thankful and greatly indebted to my supervisory committee, to Ronald L. Akers, my major advisor and Chair, Angela Gover, co Chair, Lonn Lanza -Kaduce, Karen Parker, and Ron ald Randles) for their time, knowledge, accessibility, and most of all, patience.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 8 ABSTRACT ........................................................................................................................................ 11 CHAPTER 1 INTRODUCTION ....................................................................................................................... 13 Introduction ................................................................................................................................. 13 Significance of the Study ............................................................................................................ 14 2 LITERATURE REVIEW AND THEORETICAL FRAMEWORK ....................................... 16 Definition of Dating Violence .................................................................................................... 16 Literature Review ........................................................................................................................ 17 Victim C haracteristics ......................................................................................................... 17 Demographic factors .................................................................................................... 18 Early family of origin experience ................................................................................ 2 0 Dating relationship variables ....................................................................................... 24 Pe rpetrator Characteristics .................................................................................................. 25 Early family of origin experience ................................................................................ 26 Dating relationship variables ....................................................................................... 30 Theoretical Models of Dating Violence ..................................................................................... 31 Self Control Theory ............................................................................................................. 31 Intergenerational Transmission Theory (Social Learning Theory) .................................. 33 Social Bonding (Attachment) Theory ................................................................................ 37 3 METHODOLOGY ...................................................................................................................... 39 Data Collection Pr ocedure .......................................................................................................... 39 Measur ement of Variables .......................................................................................................... 42 Dependent Variables: Perpetration and Victimization in Dating Relationships .............. 42 Independent Variables ......................................................................................................... 45 Experience of violence in the family in childhood .................................................... 45 Self -control ................................................................................................................... 46 Attachment to p arents .................................................................................................. 47 Control Va riables ................................................................................................................. 48 Data Analyses .............................................................................................................................. 50

PAGE 6

6 4 RESULTS .................................................................................................................................... 54 Korean Dating Violence ............................................................................................................. 54 Descriptive Statistics for the Korean Sample .................................................................... 54 Results of Multivariate Analyses for the Korean Sample ................................................. 57 Perpetration of physical and psychological dating violence ..................................... 57 Physical and psychological victimization by dating partners .................................... 60 Perpetration of physical and psychological dating violence : the subgroup of perpetrators only ....................................................................................................... 63 Victimization by physical and psychological dating violence : the subgroup of victims only ............................................................................................................... 66 Dating Violence by Gender in Korea ................................................................................. 70 American Dating Violence ......................................................................................................... 72 Descriptive S tatistics for the American Sample ................................................................ 72 Results of Multivariate Analysis for the American Sample ............................................. 75 Perpetration of physical and psychological dating violence ..................................... 76 Physical and psychological victimization by dating partner ..................................... 79 Perpetration of physical and psychological violence : the subgroup of perpetrators only ....................................................................................................... 82 Victimization by physical and psychological violence: the subgroup of victims only ............................................................................................................................ 85 Dating Violence by Gender in the U.S. .............................................................................. 87 Full Model Results (including U.S. and K orea) ........................................................................ 90 Descriptive Statistics for the Full Sample .......................................................................... 90 Results of Multivariate Analyses for Full Model .............................................................. 93 Physical and psychological perpetration of dating violence in the full model ......... 93 Physical and psychological victimization by dating partner in the full model ........ 95 Perpetration of physical and psychological violence among perpetrators ................ 98 Victimization by physical and psychological dating violence among victims ...... 101 Coefficient Comparison ............................................................................................................ 104 Comparison by Country .................................................................................................... 104 Comparison by Gender ...................................................................................................... 106 Gender difference in Korea ....................................................................................... 106 Gender difference in the U.S. .................................................................................... 107 5 CONCLUSION ......................................................................................................................... 135 Summary Discussion ................................................................................................................. 135 Self C ontrol Theory ........................................................................................................... 136 Direct and Vicarious Experience with Violence in the Family ...................................... 137 Attachment to Parents ........................................................................................................ 140 Control Variables ............................................................................................................... 140 Limitations ................................................................................................................................. 144 Policy implication ..................................................................................................................... 145 Future Research ......................................................................................................................... 146

PAGE 7

7 APPENDIX A SURVEY INSTRUMENT ........................................................................................................ 148 B CODE BOOK ............................................................................................................................ 153 C CORRELATION CHARTS ..................................................................................................... 157 D TABLES OF LOGISTIC REGRESSION MODELS ............................................................. 161 LIST OF REFERENCES ................................................................................................................. 173 BIOGRAPHICAL SKETCH ........................................................................................................... 182

PAGE 8

8 LIST OF TABLES Table page 4 1 Descriptive statistics for the Korean sample ...................................................................... 109 4 2 OLS regression models for perpetration of physical violence (all students) in Korea .... 110 4 3 OLS regression model for victimization by psychological abuse (all students) in Korea ..................................................................................................................................... 111 4 4 OLS regression model for perpetration of physical violence and psychological abuse (subgroup of perpetrators only) in Korea ........................................................................... 112 4 5 OLS regression model for victimization by physical violence and psychological abuse (subgroup of victims only) in Korea ........................................................................ 113 4 6 Descriptive statistics for the American sample .................................................................. 114 4 7 OLS regression models for perpetration of physical violence (all students) in the U.S. 115 4 8 OLS regression models for victimization (all students) in the U.S. ................................. 116 4 9 OLS regression model for perpetrators (subgroup of perpetrators only) in the U.S. ....... 117 4 10 OLS regression model for victimization by dating violence (subgroup of victims only) in the U.S. ................................................................................................................... 118 4 11 Descriptive statistics for full sample ................................................................................... 119 4 12 OLS regression models for perpetration of dating violence (full model) in the U.S. and Korea .............................................................................................................................. 120 4 13 OLS regression models for victimization by dating violence (full model) in the U.S. and Korea .............................................................................................................................. 121 4 14 OLS regression models for perpetration of dating violence in the U.S. and Korea (subgroup of perpetrators only) ........................................................................................... 122 4 15 OLS regression models for victimization by dating violence in the U.S. and Korea (subgroup of victims only) ................................................................................................... 123 4 16 OLS regression models for perpetration of psychological abuse by gender in Korea (overall model) ..................................................................................................................... 124 4 17 OLS regression models for victimization by psychological abuse by gender in Korea (overall model) ..................................................................................................................... 125

PAGE 9

9 4 18 OLS regression models for victimization by psychological abuse by gender in Korea (subgroup of victims only) ................................................................................................... 126 4 19 OLS regression models for perpetration of physical violence by gender in the U.S. (overall model) ..................................................................................................................... 127 4 20 OLS regression models for perpetration of psychological abuse by gender in the U.S. (overall model) ..................................................................................................................... 128 4 21 OLS regression models for victimization by psychological abuse by gender in the U.S. (overall model) ............................................................................................................. 129 4 22 OLS regression models for perpetration of physical violence by gender in the U.S. (subgroup of perpetrators only) ........................................................................................... 130 4 23 OLS regression models for perpetration of psychological abuse by gender in the U.S. (subgroup of perpetrators only) ........................................................................................... 131 4 24 Significance of d ifference in OLS regression coefficients for male and female models in Korea .................................................................................................................... 132 4 25 Significance of d ifference in OLS regression coefficients for male and female models in the U.S. ................................................................................................................ 133 4 26 Significance of difference in OLS regression coefficien ts for Korean and American models ................................................................................................................................... 134 D 1 Logistic regression model for perpetration of physical violence in Korea ....................... 161 D 2 Logistic regression model for perpetration of psychological abuse in Korea .................. 162 D 3 Logistic regression model for victimization by physical violence in Korea .................... 163 D 4 Logistic regression model for victimization by psychological abuse in Korea .............. 164 D 5 Logistic regression model for perpetration of physical violence in the U.S. ................... 165 D 6 Logistic reg ression model for perpetration of psychological abuse in the U.S. ............... 166 D 7 Logistic regression model for victimization by physical violence in the U.S. ................. 167 D 8 Logistic regression model for victimization by psychological abuse in the U.S. ............ 168 D 9 Logistic regression result of full model -perpetration of physical violence ...................... 169 D 10 Logistic regression result of full model -perpetration of psychological abuse ................. 170 D 11 Logistic regression result of full model -victimization by physical violence .................... 171

PAGE 10

10 D 12 Logistic regression result of full model -victimization by psychological abuse ............... 172

PAGE 11

11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FAMILY EXPERIENCE, ATTACHMENT TO PARETNS, AND LOW SELF CONTROL IN DATING VIOLENCE IN THE U.S. AND KOREA By Mi Rang Park May 2009 Chair: Ronald Akers Cochair: Angela Gover Major: Criminology, Law and Society Although many studies have been done about dating violence, only a few researchers have developed theoretical conceptualizations of factors such as these in dating violence (Sellers, 2005; Follingstad et al, 1999; Riggs and OLeary, 1989; Luthra and Gidycz, 2006). This dissertation moves beyond these to draw on three prominent general theori es of crime and deviance for its theoretical framework to understand the effects of early family experience with violence (social learning), current attachment to parents (social bonding), and self -control. The purpose of this study is to link theoretica lly relevant variables that are linked in some way to the respondents family of origin that may account for the likelihood of perpetration of or victimization by dating violence in two different cultures. In using the same survey instrument for two count ries, this study tested whether theoretical explanations developed in Western society could be applied to Eastern society. A sample of 1,400 young adults attending a university in Korea and a sample of 1,500 young adults attending a university in America w ere used in this dissertation. Findings indicate that experience of child abuse (social learning), low self -control (self -control), serious dating relationship and risky sexual behavior have significant effects on both the perpetration and victimization of dating violence in both countries. Self -

PAGE 12

12 control has stronger effects in the American, and being the victim of violence in the home has more significant effects in the Korean sample.

PAGE 13

13 CHAPTER 1 INTRODUCTION Intro duction The research reported in this dissertation investigated dating violence victimization and perpetration among sample university students in the United States and Korea. Dating violence includes: psychological or emotional violence, such as controlli ng behaviors or excessive expressions of jealousy; physical violence, such as hitting or punching; and sexual violence such as nonconsensual sexual activity and rape. While both females and males may suffer dating violence, females were more likely to be i njured, more likely to be sexually assaulted, and more likely to suffer emotionally than males. The Center of Disease Control and Prevention (CDC) has reported that one in 11 adolescents report physical violence and one in five adolescents report emotiona l abuse. A separate research study indicated that approximately one in five female students report being physically or sexually abused by a dating partner (Silverman, Raj, Mucci, & Hathaway, 2001). In a study conducted in North Carolina 66.7 % of the respondents reported sexual or physical violence with a partner while in high school (Smith, White, & Holland, 2003). In another study, over 70% of female students reported that they had experienced one or more incidents of unwanted sexual activity (Jackson, Cram, & Seymour, 2000). Physical and sexual dating violence against adolescent girls has been associated with increased risk of substance use, unhealthy weight control behaviors, risky sexual behaviors, pregnancy, and suicide attempts (Silverman et al., 2001). Moreover because t he students either did not tell anyone about the sexual violence, or only told a friend (Jackson et al. 2000) we can expect that an important aspect of dating violence is its hidden or veiled nature

PAGE 14

14 While preliminary re search in the area of dating violence is promising and certainly a step in the right direction, the study of dating violence and aggression is limited by little theoretical development (Sellers, Cochran & Branch, 2005). This dissertation will focus on the relationship between early family of origin experience, relationship with parents, issues of self -control, and both the perpetration of, and victimization by, dating violence in Korea and the U.S. Because m ale perpetrators of dating violence are usually y ounger than 25 years old (Hunnicutt, 1998) the college student is a good subject to study dating violence. T his project will investigate how these variables are linked with dating violence perpetration and victimization. Although the data do not allow ad equate direct tests of social learning and social bonding theories, they provide theoretical frameworks for understanding effects of variables relating to the family of origin of the respondents. The data do allow a direct test of self -control theory that has been suggested in the literature as an explanation of dating violence (Sellers et al. 2005). Significance of the Study Violence has long been an area of interest for social scientists and practitioners (Lewis & Premouw, 2001). Recently, several studi es have been conducted on dating violence behavior among young adults. These reports tend to be descriptive studies without attention to theoretical issues. This is true also of international studies of dating violence. One important contribution of this d issertation is to apply general criminological theories to explain the perpetration and victimization by dating violence. M ost often applied in the literature are either the psychological or the youth development theories. This project will bring explanati on of perpetration and victimization by dating violence closer to recognized criminological perspectives. Its other main contribution is to apply these theories to cross -national comparisons; specifically, between student populations in the United States a nd Korea. One of the main advantages of a cross -

PAGE 15

15 national comparative study is the ability to investigate the extent to which general explanations of crime developed primarily in Western society can be generalized to behaviors in other societies. The theo retical issue is what explanatory variables account for or predict variations in this form of violence. A considerable amount of research exists on the causes and consequences of intimate violence in the United States, with a major focus on commission of violence by males and victimization by females. Not much of this research focuses on victimization by males or on comparing male with female victimization and very little of it involves comparison with victimization in other societies. Therefore, another c ontribution of the dissertation is that it will fill lacunae not only in cross cultural comparisons but also in gender comparisons. In brief, the dissertation addresses the questions of which set of factors might be reflective of how three major criminolog ical theories: those of self -control, social learning, and social bonding, best account for dating violence, both for victim and offender. Further, are these factors the same or different in two different societies for both men and women? This dissertati on begins in Chapter 2 to detail the three theoretical perspectives that are of interest in this study. Chapter 3 discusses the data and methodology. Chapte r 4 provides the result of the multivariate analysis. Chapter 5 contains a conclusion and discussion on direction for future research.

PAGE 16

16 CHAPTER 2 LITERATURE REVIEW AN D THEORETICAL FRAMEW ORK Definition of D ating V iolence There is a fairly wide variation in estimates of frequency of intimate violence and that is in part the result of several definition and measurement issues in the literature. Although the majority of research has focused on physical violence, other studies have examined psychological/emotional abuse, threats of violence, verbal abuse, and sexual violence (Gover et al 2008). Studies that only examine one type of abuse or use a narrow definition of abuse report lower estimates than studies that examine several types of abuse according to a broad definition. Some studies report lifetime prevalence estimates and other studies report dat ing violence experienced during the prior twelve months. Lack of consensus of on operational definition of dating violence is a difficult issue, complicating the investigation of courtship battering. Dating violence is a vague term that can include threate ning communication, verbal abuse, or physical aggression. Due to such ambiguity, prevalence rates widely fluctuate. R esearchers have estimate d that between 9% and 87% of high school and college students are involved in a violent dating relationship (Harned 2002; Watson, Cascardi, Avedry-Leaf, & O L eary, 2001). Harned (2002) reported over 80% of men and women were victim s of psychological abuse and about 20% of men and women were victims of physical violence from a partner. However, Cercone, Beach, an d Aria s (2004) reported less than 30% of men and women were exposed to psychological abuse by their partner, and less than 20% of men and women were victims of physical violence from a partner. The definition of dating violence used in this dissertation is based on the U.S. Department of Justice definition of dating violence as the perpetration or threat of an act of violence by at least 1 member of an unmarried couple on the other member within the context of dating or

PAGE 17

17 courtship (Ramisetty -Mikler, Goebert, Nis himura, & Caetano, 2006, p.423). It is also derived from The Centers of Disease Control and Prevention (CDC) definition as encompassing any forms of sexual assault, physical violence, and verbal or emotional abuse within the dating or courtship relations hip. In this dissertation dating violence is divided into physical dating violence, including physical violence and coercive sexual behavior, and psychological dating violence, including verbal abuse and emotional abuse. Literature Review A number of vari ables have been postulated in the literature as predisposing, moderating, or mediating factors related to dating aggression, including gender (Monson & Lanhinrichsen Rohling, 2002; Gover et al., 2008); exposure to childhood violence (Follinstad, Bradely, L aughlin & Burke, 1999; Gover et al., 2008; Shook et al., 2000); previous victimization (Banyard et al., 2006 ; Hickman et al., 2004; Smith, White & Holland, 2003); attitudes about interpersonal aggression (Foshee et al., 2005); alcohol use (Gover, 2004; Gover et al., 2008; OKeefe, 1997; Shook et al., 2000); and low self -esteem and depression (Foshee et al., 2004; Marshall & Rose, 1990). Much empirical research exists related to dating violence ; however, most studies have reported result s without separation of victimization and perpetration. Frequently result s were reported for just victimization or perpetration of dating violence not both Although v ariables impacting victimization are often similar to those associated with perpetration different factors for perpetration and victimization have been found in empirical research. Empirical researches will be reviewed separately for victimization and perpetration of dating violence Victim C haracteristics In this section, I will review the victim -specific variables to provide valuable information toward a better understanding of victimology and possible risk factors. In this section,

PAGE 18

18 demographic information, early family of origin experience and dating relationship variables will be shown as victimization factors for dating violence. Demographic factors Demographic information helps identify individuals experiencing the phenomena in question. Demographics generally refer to gender, socioeconomic status, age, race, and area of resid ency. However, demographic variables have not been investigated separately for victims and perpetrators. As the result, information will be provided for individuals experiencing dating violence, either by initiation or victimization. An examination by vic tims gender yields vastly different perpetration rates of dating violence (Arias et al., 1987; Clark et al., 1994; Lane & Gwartney Gibbs, 1985; Marshall & Rose, 1987; OKeefe et al, 1986; Riggs et al., 1990). Early research on dating violence for adolesc ents has tended to report that females were more likely to be victims (Roscoe & Kelsey, 1986). For example, Bergman (1992)s study reported that females are more often victimized by sexual and physical violence by their dating partner than are males. Simil ar studies have also shown ahigher rate of female victimization (Arriaga & Foshee, 2004; Harned, 2002; Mar s hall & Rose, 1988). More recent research, however, has reported equal or higher victimization rates for males. For example, Alexander, Moore and Ale xander (1991) indicated males reported a higher rate of victimization by verbal and physical violence in dating violence than did females. In addition, other studies have found a higher level of male victimization in other countries. Schiff and Zeira (2005) explored experiences of dating violence in Israel using convenience sampling. They found a high rate of victimization by physical violence among boys. Lysova (2007) found a similar result in Russia: the victimization rate of males by female partners was greater than female victimization by male partners for both minor and severe forms of physical violence. In

PAGE 19

19 terms of attitude toward violence perpetration, Jackson (1999) and OKeefe (1997) found that females may not stigmatize their perpetration of datin g violence as much as males who reported violence perpetration. As a result, higher rates of male victimization may be reflected by the fact that females feel more comfortable reporting their violent behavior than males do (Gover et al., 2008). Age has tr aditionally not been found to be a significant predictor of dating violence (Arias et al., 1987; Marshall & Ro se, 1987; Reuterman & Burcky, 1989), although this may be a function of limited empirical investigation (Stets & Henderson, 1991). Much of the da ting violence research excludes ethnic minorities due to small sample sizes (Sugarman & Hotaling, 1989). As a result, an accurate representation of dating violence in nonwhite groups is speculative. Several studies have directly attempted to address this question, with mixed results. Some researchers demonstrated higher rates of violence among African Americans when compared to Caucasians (Makepeace, 1987; OKeefe, 1997; OKeefe et al., 1986). Other researchers, by contrast, suggested that violence may be more prevalent among Caucasians (Lane & Gwartney Gibbs, 1985), while other reports indicated there may be no differences in dating violence as a function of race (Clark et al., 1994; White & Koss, 1991). Family structure also has been tested as a source o f demographic information. Previous studies sought data regarding respondents family structure in which they were raised. Usually the respondents selected their response from choices such as lived with two biological parents, lived with single parent, lived with stepparents, and lived with others, such as adoptive parents or other relatives. This area of investigation provides discrepant findings. Although previous findings did not support this proposition (Billingham & Gilbert, 1990), victims of d ating violence appear to be more likely to grow up in families marked by parental divorce or

PAGE 20

20 separation than to have grown up with both parents (Billingham & Notebaert, 1993). In a similar vein, Gover (2004) tested the impact of family structure on risk fo r victimization. She found that being raised by two parents decreased the odds of victimization among high school students in South Carolina (Gover, 2004). Similarly, other research has found that those students who were raised by a single parent were more likely to be a victim by dating partners (Chase, Treboux, & O Leary, 2002; Makepeace, 1987; Malik et al., 1997). However, Billingham and Gilbert (1990) reported a trend in their data. Males from divorced families reported levels of violence for themselve s and their partners which approached significance. It is also more likely that victims of dating aggression had experienced periodic or total absence by either their mother or father figures (Makepeace, 1987; OKeefe et al., 1986). Early family of origin experience Family of origin experience refers to the significant caretakers and siblings that a person grew up with, or the first social group a person belongs to, which is often a persons biological family or an adoptive family (Bandura, 1973). Many st udies have supported the relationship between family of origin experience and dating violence in later life. Variables, such as experiencing child abuse, witnessing inter -parental violence, self -control, and attachment to parents, have been investigated re garding their effect on subsequent victimization. Researchers have traditionally believed that experiencing physical abuse as a child increased the risk of later aggressive behaviors, based on social learning theory (Bandura, 1973). Social learning theory proposes that individuals acquire novel behaviors and explained their behavioral repertories by observing the behavior of others. This model suggests that observational learning is equally as important in shaping new behavior as is classical and operant co nditioning (Banduara 1973). Individuals watch another person and then use imitation to perform novel behaviors.

PAGE 21

21 Many scholars posit that a pattern exists between experiencing physical abuse as a child and being the recipient of dating violence. Experienc e of violence in childhood is one of the most frequently studied risk factors for general domestic violence including dating violence (Langhinrichsen -Rohling, Hankla & Stormberg, 2004). Makepeace (1987) found that victims of dating violence were more like ly to report exceedingly harsh discipline by parents. OKeefe et al., (1986) found that nearly half of their sample who was exposed to childhood physical abuse had experienced dating violence. Similarly, Smith and Williams (1992) reported that subjects who were raised by abusive parents had a significantly higher occurrence of dating violence. Two studies uniquely attempted to look at physical abuse separately for perpetration and victims. OKeefe (1998) demonstrated that a history of childhood physical abu se significantly distinguished female victims of dating violence from individuals in nonviolent relationships. It is important to note, however, that experiencing child abuse was a risk factor for female, but not male victims. Gover et al ( 2008) supported the gender specific effect of child abuse experience on victimization by dating violence. Childhood abuse was associated with the likelihood of dating violence victimization for females (Gover et al., 2008; Marshall & Rose, 1988). However, this gender spe cific effect was not found in victimization by psychological abuse (Gover et al., 2008). Coffey Leitenberg, Henning, Bennett, and Jankowski (1996) demonstrated that the victims of dating violence were significantly more likely to had experienced child abu se, as compared to nonviolent controls. Witnessing inter -parental violence has been proposed to be a risk maker for subsequent victimization. Several studies found evidence for a modeling conceptualization of dating aggression (Coffey et al., 1996; OKeef e et al., 1986). In contrast, Marshall and Rose (1988) did not demonstrate support for the intergenerational transmission of violence. This research

PAGE 22

22 indicated that the relation between dating violence and child abuse was significantly strengthened when vic tims also witnessed inter -parental violence (OKeefe, 1998). Moreover, this result appears to have more sex specific effects. Gover et al., (2008) found mixed results compared to previous patterns. According to Gover and her colleagues, there was significa nt correlation between witnessing inter -parental violence and victimization by physical violence in dating violence. Moreover, t his effect is greater for female students. However, they failed to find a significant impact of witnessing of parental violence on victimization by psychological abuse. International studies also have reported a significant impact of the experience of child abuse on victimization by dating violence. A New Zealand study by Fergusson, Boden and Horwood (2005) and a Canadian study by Kwong Bartholomew Henderson, and Trinke (2003) reported significant impact of learned abusive behavior on victimization. Fergusson et al. (2005) examined whether exposure to inter parental violence in childhood predicts subsequent involvement in inter -partner violence using a birth cohort of over 1000 New Zealand young adults. They found that exposure to inter -parental violence in childhood increased victimization by psychological abuse later. Although they failed to find the association between victimization and physical violence, previous research conducted in Canada by Kwong et al., (2003) found that learned abusive behaviors including physical violence and psychological abuse are transmitted across generations in both forms of victimization and perpetration. Another variable related to early family of origin experience is self -control. Self -control is an individual trait that is developed in early childhood. Because low self -control is the result of ineffective socialization, especially due to ineffective parenting (Gottefredson & Hirschi, 1990), it is strongly related to variables of early family of origin experience. Few studies have tested the application of this self -control to victimization by dating violence. Schreck, Steward and Fisher

PAGE 23

23 (2 006) studied the relationship between victimization and self -control. They found that low self control measured at an earlier time is associated with later victimization and later risky lifestyles. Although the study was not directly related to dating viol ence victimization, it is an example suggesting the relationship between self -control formed in childhood and later victimization in life. Two recent studies further drew the relationship between self -control and victimization by dating violence. Gover et al., (2008) found that higher levels of self -control reduced the likelihood of victimization by physical violence and psychological violence among college students. In addition, Kerley, Xu and Sirisunyaluck (2008) supported the impact of self -control on vi ctimization by intimate partner violence among married females in Bangkok, Thailand. However, their finding demonstrated that only certain dimensions of self -control, such as impulsivity and risk taking, were significant predictors of intimate partner violence in Thailand. In addition to self -control, attachment is an important factor related to early family of origin experience. Bowlby (1973) suggested that children form mental representation or prototypes of relationships based on their experiences with primary caregivers during childhood. According to his suggestion, the quality of the child -parent attachment relationship affects later psychological functioning (Bowlby, 1973; Sternberg, Lamb, Guterman, Abbott & Dawud Nouris, 2005). In a similar vein, C leve land et al. (2003) studied the relationship between attachment to parents and victimization by dating violence. They reported that females who were attached to their mothers were less likely to experience dating violence victimization by their partner. Similarly, Gover et al., (2008) found that maternal attachment reduced the likelihood of physical violence victimization for males, but not for females. However, Gover et al., (2008) failed to establish the impact of parental attachment on victimization b y psychological abuse.

PAGE 24

24 Dating relationship variables Dating relationship factors include the number of dating partners, the seriousness of the dating relationship, and the length of time of the dating relationship. Empirical studies have shown a positive relationship between those dating relationship factors and the incidence of victimization by dating violence. The association functions somewhat differently by gender. Commonly, previous research has explained relati onship factors as one measure of sexual victimization, in addition to victimization by physical violence and psychological abuse. The depth or seriousness of a dating relationship may be associated with abuse: as the level of the relationships perceived c ommitment deepens, both parties will also have more invested, emotions may become more intense, and the opportunities for conflict may increase (Cleveland et al. 2003). Studies indicate that dating violence tends to occur more often in a serious datin g relationship (Cate et al., 1982; Henton et al., 1983; Sigelman et al., 1984). Harned (2002) found that a greater number of casual dating partners as well as a history of sexual aggression toward dating partners were predictive of an increased risk of sex ual victimization by dating partners among both men and women. In addition, psychological aggression toward dating partners was associated with an increased risk of sexual victimization of dating partners for women. However, Harned (2002) indicated that da ting behaviors were significant predictors only of sexual and psychological victimization. Specifically, sexual victimization was associated with engaging in more casual dating relationships, whereas a greater number of serious dating partners predicted ps ychological victimization. These findings are consistent with previous research indicating that sexual victimization is more common when the dating partner is someone with whom the victim has had fewer dates (Muehlendahrd & Linton, 1987), whereas psycholog ical victimization is associated with an increased duration of the dating relationship (Sharp & Taylor, 1999).

PAGE 25

25 Perpetrator C haracteristics Identification of predictor variables or risk makers related to the perpetration of dating violence is impor tant I n this section, I will review the perpetrator -specific characteristics associated with perpetration of violence The variables will be presented similarly to those presented for victims, including demographics, early family of origin experience, and dating relationship variables. Because demographic information is similar to that provided for victims, only a cursory review of important demographic information will be provided. Previous research indicated that perpetration rates vary by gender (Arias et al., 1987; Clark et al., 1994; Lane & Gwartney Gibbs, 1985; Marshall & Rose, 1987; OKeefe et al., 1986; Riggs et al., 1990). Early research on adolescent dating violence suggested that males were more likely to be perpetrators of dating violence (Roscoe & Kelsey, 1986). For example, Bergman (1992) reported that males were significantly were more likely than females to be perpetrators of sexual and physical violence. Other studies have reported similar results in terms of gender difference on dating violenc e (Arriaga & Foshee, 2004; Marshall & Rose, 1988). Recent studies, however, have found contrary results: females were more likely to be perpetrators of physical and psychological abuse (Cercone et al., 2005; Foshee et al., 1996; Straus, 2004). Gover et al. (2008) supported the finding of a higher rate of female perpetration using a sample of college students. Gover and her colleagues found that males were significantly less likely than females to perpetrate physical violence, and specifically, that being m ale decreased the odds of physical violence perpetration by 50%. This gender effect was significant in the perpetration of psychological abuse as well. Specifically, male students were significantly less likely than female students to perpetrate psychologi cal abuse, and being male decreased the odds of perpetration of psychological abuse by 35% (Gover et al., 2008).

PAGE 26

26 There has been equivocal support for the relations between age and dating violence (Arias et al., 1987; Marshall & Rose, 1987; Reuterman & Bur cky, 1989). Similarly, support for the association of ethnic minority status and dating violence has been inconclusive (Clark et al., 1994; Lane & Gwartney Gibbs, 1985; Makepeace, 1987; OKeefe, 1997; White & Koss, 1991). Akin to previous victimization stu dies that located a relationship between family structure and victimization and dating violence, the influence of family structure on perpetration of dating violence has also been supported in the literature. Research provides only marginal support for an association between parental divorce and the perpetration of dating violence. Tontodonato and Crew (1992) found an increased possibility of perpetration of dating violence in males with divorced parents. Similarly, Billingham and Notebaert (1993) demonstra ted that individuals from divorced families employ dating violence as a conflict resolution strategy more often than do individuals from intact families. Other findings, however, have demonstrated only a trend toward significance in their data. Males from divorced families reported initiating more violence in dating relationships than males from intact families (Billingham & Gilbert, 1990). Early family of origin experience Factors of early family of origin experience such as child abuse, witnessing inte r -parental violence, low self -control, and parental attachment are all believed to play a role in the perpetration of dating violence. More scholars have focused on the perpetration of dating violence and its links to the experience of family violence in c hildhood than on victimization studies of dating violence. The experience of family violence usually refers to direct experience of child abuse and witnessing inter -parental violence. Bernard and Bernard (1983) indicated a direct mirroring, in that the ex act type of aggression observed or experienced in the family was perpetrated in dating violence.

PAGE 27

27 Experience of violence in childhood is one of the most frequently studied risk factors for general future violent behavior, including dating vi olence (Langhin richsen -Rohling et al., 2004). W hile other researchers suggest there is no significant relation between child abuse and dating aggression ( Carr & VanDeusen, 2002; Foo & Margolin, 1995; OKeefe, 1997) ,most of researchers have documented a significant relat ion between experiencing child abuse and perpetrating dating violence (Laner & Thompson, 1982; Marshall & Rose 1988). Parents who were physically punished as children showed a greater rate of frequent child abuse and spouse abuse (Straus & Gelles, 1986, 1990; Straus, Gelles & Steinmetz, 1980). Widom (1989) found that high proportion of those who had been physically, sexually abused or neglected as children were arrested as adults for sex crimes. Abused or neglected males were almost twice as likely to be a rrested for rape 20 years later. Similarly, victims of physical child abuse had the highest level of arrest for violent criminal behavior (Widom, 1989). Other studies have also supported this link between child abuse and violent behavior later in life (Dog e et al, 1990; Lisak et al., 1996; Marshall & Rose 1988). While males and females who experienced child abuse in childhood were significantly more likely to perpetrate dating violence, many previous studies found a gender specific effect of child abuse on the perpetration of dating violence. Specifically, the impact of experience of child abuse is stronger for males. Alexander, Moore and Alexander (1991) found that there was a significant relationship between being abused as a child and perpetrating abus e only for those men who had received severe abuse from their fathers. Marshall and Rose (1988) found that being abused as a child was related to dating violence victimization and perpetration among males, but not for females. Similarly, Wekerle and Wolfe (1998) reported that the effect of childhood maltreatment applied to male perpetrators of physical violence. However there are

PAGE 28

28 different results that indicated reversed relationship between child abuse and gender specific effect on dating violence. Folle tte and Alexander (1992) and O Keefe (1998 ) have found childhood abuse to be a risk factor for females initiating dating violence, but not so for males (Follette & A lexander, 1992; OKeefe, 1998) Although discussion for gender specific effect is inconclus ive, overall empirical support has been found for the influence of experience of child abuse and perpetration of dating violence. Moreover, Marshall and Rose (1988) argued that the influence of direct experience of child abuse is greater than witnessing of indirect experience of violence in the family. Witnessing inter -parental violence is also believed to be associated with the perpetration of dating violence. Similar to the effects of childhood abuse, empirical support is indeterminate. There is, however some evidence of a relation between witnessing of parents violence and the perpetration of dating violen ce ( Bernard & Bernard, 1983; GwartneyGibbs et al., 1987; Laner & Thompson, 1982; OKeefe, 1997; Ronfeldt, Kimerling, & Arias, 1998). Gover et al. (2008) found that college students who witnessed maternally perpetrated violence during childhood were significantly more likely to perpetrate psychological abuse in dating violence. Specifically, respondents who witnessed their mother hit their father incre ased the odds of perpetration of psychological abuse by almost 60% (Gover et al., 2008). Still other data indicated the same trend for female -only perpetration of dating violence (Follette & Alexander, 1992; Riggs & OLeary, 1996). More specifically, OKeefe (1998) found that for males who witnessed significant inter parental violence, accepting the use of violence in relationships distinguished perpetrators from nonviolent males. Similarly, for females who witnessed high levels of inter parental violence, the additional experience of childhood abuse distinguished perpetrators from nonviolent females.

PAGE 29

29 Several researchers, however, have found no significant correlation between witnessing parental aggression and the expression of dating violence (Riggs et al., 1990; Tontodonato & Crew, 1992). Maxwell and Maxwell (2003) found that child-directed and child -witnessed violence in the family influenced later adolescent violence among study groups in the Philippines. In their study, the effects of family violence me asured on rates of aggression do not vary by gender, suggesting that forms of family violence have similar negative consequences for both boys and girls (Maxwell & Maxwell, 2003). Another family related factor is self -control. Self -control is an individua l trait that is developed in early childhood. Low self -control is the result of ineffective socialization, especially due to ineffective parenting (Gottefredson & Hirschi, 1990). Sellers (1999) found low self -control to have a modest but significant influe nce on physical violence perpetration among college students. Using a sample of 985 college students, she found that self -control explained 10% of the variation in courtship violence, and when combined with measures of opportunity and rewards, explained ab out 17% of the variation (Sellers, 1999). Clapple and Hope (2003) reported that low self -control influenced significantly on perpetration of physical dating violence among high school students. A recent study conducted by Gover et al., (2008) supported the influence of self -control on perpetration of physical violence and psychological abuse among college students. Higher scores of self -control reduced the odds of perpetration of dating violence including both physical violence and psychological violence a mong college students (Gover et al., 2008). Kerley et al. (2008) tested the impact of low self -control on aggression and victimization by intimate partner violence among married females in Thailand. Like previous results of victimization, self -control mode rately influenced the perpetration of dating violence.

PAGE 30

30 Usually the self -control proposed by Gottfredson and Hirschi (1990) reflects uni -dimensionality however, Kerley et al. (2008) tested the separate effects of each of the six dimensions of self control. As supporting the multi -dimensionality of self -control, they found that low self -control partially explained the aggression of dating violence. Impulsivity was a robust predictor of both types of intimate partner aggression. Risk -taking was a significant predictor only of physical forms of aggression. Low frustration tolerance was a predictor of both types of intimate partner aggression, but neither type of victimization (Kerley et al., 2008). In addition to self -control, attachment is an important facto r related to early family of origin experience. Clapple and Hope (2003) reported that high school students who scored higher attachment to their parents were less likely to perpetrate dating violence. Specifically, Gover et al. (2008) found the gender effe ct of parental attachment on perpetration of dating violence. Gover et al. (2008) reported that father attachment decrease the likelihood of perpetration of physical violence and psychological abuse for college students. In addition, they found that matern al attachment significantly reduced the likelihood of physical violence perpetration for male students. And strong attachment to father reduced the likelihood of perpetration of psychological abuse for female students. Dating r elationship variables Empirical studies have reported the positive relationship between dating relationship factors and perpetration of dating violence. OKeefe (1997) found that the seriousness of the relationship, measured with a single item asking participants to rate the se riousness of their relationship from 1 (casual dating) to 7 (someone you are engaged to marry), was a significant predictor of perpetrating abuse for females, but not for males. Gover et al., (2008) found that having a steady or exclusive dating partner in creased the risk of physical dating violence perpetration. Consistent with physical violence, students engagement in a steady or exclusive

PAGE 31

31 dating relationship increased the likelihood of perpetration of psychological violence (Gover et al., 2008). The ser iousness of the dating relationship may be associated with abuse because as seriousness increases, both parties have more invested in the relationship, emotions may become more intense, and the opportunities for conflict may increase ( Cleveland et al. 2003). A lthough it did not directly test the dating violence and dating relationship, Betheke and Dejoy (1993) examined the acceptability of dating violence. The result indicated that experimental subjects were more tolerant of violent behavior when the depic ted relationship was serious and when the perpetrator was female (Detheke & Dejoy, 1993) Theoretical Models of Dating Violence However, only a few researchers have developed theoretical conceptualizations of factors such as these in dating violence (Sell ers, 2005; Follingstad et al, 1999; Riggs & OLeary, 1989; Luthra & Gidycz, 2006). This dissertation moves beyond these to draw on three prominent general theories of crime and deviance for its theoretical framework: social learning, self -control and socia l bonding theory. Of these the perspective that is best represented and measured in the data for the dissertation is self -control theory. Self -Control Theory According to Gottfredson and Hirschi (1990), crime is a result of a single underlying individual characteristic that they term criminal propensity or low self -control. They argue that low self -control is the result of ineffective child rearing, and is essentially formed by age 8 10. In order to instill self -control in a child a parent must supervise t he child, recognize acts of deviant behavior, and punish the child for those acts. If the parent is ineffective in these areas, the child will develop low self -control, and will have a propensity to commit crime. Low self control is characterized by impuls ivity, insensitivity, risk taking or thrill -seeking, physicality, preference for easy rather than complex tasks or short -sightedness, and the preference for

PAGE 32

32 nonverbal rather than verbal tasks1. Gottfredson and Hirschi (1990) argue that, once formed in chil dhood, self -control is relatively stable throughout the life course, and the cross individual difference in self -control does not change. Someone with low self -control is more likely than one with high self -control to commit crime given the opportunity thr oughout life. Thus, although current self control may be measured at a later time, the assumption of the theory is that ones self -control was formed in childhood in the family of origin. A lthough Gottfredson and Hirschi (1990) developed their general the ory to specifically explain crime and analogous behavior they stated self -control can also explain l behavior such as accidents as well. Further, we know that the commission of crime and victimization by crime are related. Also, self -control theory expe cts versatility and a variety of behavior Since low self control is expected to account for commission of criminal violence, then it should also be at least somewhat related to being the victim of violence. Then, this theory may also be useful for explaining victimization Indeed, some studies have taken that approach and found support for hypothesized link s between low self -control and an increase in crime victimization risk (Holtfreter, Reisig, & Pratt 2008; Schreck, 1999; Schreck et al, 2002; Schreck et al., 2006; Stewart et al., 2004 ). Based on the inferences about victimization that can be made from the logic of self -control theory and the body of previous research that has tested low self -control explanations of crime victimization sel f-control is expected to explain both perpetration and victimization of dating violence. Few studies have tested the application of this theory to dating violence. Using data from a national probability sample of respondents, Avakame (1998 a ) examined the role that self -control plays in the processes by which violence is transmitted through generations. Self -control was 1 The revised version of self control theory by Hirschi (2004, p. 543) defined it as the tendency to consider the full range of potential costs of a particular act

PAGE 33

33 found to mediate the relationship between witnessing violence between parents and intimate partner violence among males and females. Using these same data, Avakame (1998 b ) examined the relationship between violence in the family of origin and perpetration of psychological aggression. Support was not found for low self -control as a mediating variable in the relationship between violence in the family of origin and adulthood psychological aggression among males. Two other studies, however, reported that low self -control plays a role in the intergenerational transmission of violence. Sellers (1999) found low self -control to have a modest but significant influence on physical violence perpetration among college students. Clapple and Hope (2003) reported that low self -control was found to be a significant predictor of physical dating violence perpetration among high -school students. Kerley et al. (2008) tested the impact of low self control on perpetration of and victimization by intimate partner violence in Thailand. Although their results did not account for the behavior of dating violence directly, they found that some parts of low self -control theory contributed to explain victimization and perpetration of intimate partner violence of married females (Kerley et al., 2008). Thus, there are some empirical as well as theoretical grounds for application of self -control theory to dating violence victimization and perpetration. Intergenerational Transmission Theory (Social Learning Theory) Researchers have traditionally believed that experiencing physical abuse as a child increased the risk of later aggressive behaviors, based on social le arning theory (Bandura, 1973). Social learning theory proposes that individuals acquire novel behaviors and explained their behavioral repertories by observing anothers behavior. This model suggests that observational learning is equally as important in s haping new behavior as is classical and operant conditioning (Banduara, 1973). Individuals watch another person and then use imitation to perform novel behaviors. Bandura also asserts that individuals can learn new behaviors without performance

PAGE 34

34 and without reinforcement. Simply stated, an observer may copy a models behavior long after he or she saw the action performed and even though no immediate reinforcement was earned by the model or the observer (Lewis & Fremouw, 2002). A number of studies have measur ed exposure to violence during childhood as witnessing violence between parents and/or as being the victim of violence by a parent or guardian. Most studies have used the Conflict Tactics Scale (CTS) or a modified version of the CTS to measure exposure to childhood physical violence (Jankowski, Leitenberg, Henning, & Coffey, 1999; Marshall & Rose, 1988). A number of recent studies have examined the link between witnessing inter -parental violence during childhood and experiencing violence. Early work by Bernard and Bernard (1983) indicated a direct mirroring, in that the exact type of aggression observed or experienced in the family was perpetrated in dating relationships. However, the study of intergenerational transmission of violence via exposure to violence in the family of origin has yielded inconclusive results. It appears that growing up in a violent home does not adequately predict who will perpetrate or experience violence in intimate, adult relationships. In fact, Kaufman and Zigler (1987) report tha t only about 30% of those who witness interpersonal violence or experience violence at the hand of their parents also perpetrate violence as adults. Of course this is much higher tha n of those who had not experience with violence in the family of origin, a nd in fact is one of the stronger predictors of later violence. Several recent studies found that men who observed inter -parental violence were more likely to behave aggressively toward their girlfriends (Choice, Lamke & Littman, 1995; Riggs, OLeary, & Breslin, 1990; Straus, 1992; Straus Kaufman, & Kantor, 1994). In four studies, female participants were more likely to aggress against their partners if they witnessed inter parental violence in the family. Using a telephone survey of 1,249 Canadian adult s, Kwong,

PAGE 35

35 Bartholomew, Henderson and Trinke (2003) supported a general social learning model of relationship violence. According to their findings, witnessing parents violence, including physical and psychological abuse in the family, was predictive of re lationship abuse later. Avakame (1998) discovered that a fathers violence is more likely to be linked with violence in the current relationship than a mothers violence. Kwong et al. (2003) explored whether boys had learned their fathers violence pattern against a partner. The link between experiencing parental violence as a victim during childhood and adult, dating violence has been similarly reported. Straus and colleagues found strong support for the intergenerational cycle of violence (Straus & Gelle s, 1986, 1990; Straus, Gelles, & Steinmetz, 1980). Children who were repeatedly abused by their parents also repeatedly and severely assaulted a sibling. Parents who were physically punished as children showed a greater rate of frequent child abuse and spouse abuse. Child abuse experience is another example of learning process. In one study there was a significant relationship between being abused as a child and perpetrating abuse only for those men who had received severe abuse from their fathers (Alexande r et al. 1991). Widom (1989) found that higher percentage of those who had been physically or sexually abused or neglected as children were arrested as adults for sex crimes. Abused or neglected males were almost twice as likely to be arrested for rape 20 years later. Similarly, victims of physical child abuse had the highest level of arrests for violent criminal behavior. Other studies have supported this link between child abuse and violence later in life (Doge et al., 1990; Lisak et al., 1996, Marshall & Rose, 1988). A lthough the majority of previous research has supported the positive relationship between child abuse and future violence, Carr and VanDeusen (2002) did not find a significant

PAGE 36

36 relationship between physical violence experienced in childhood and dating violence perpetration. These studies have been seen as supporting the hypothesis of intergenerational transmission of violence as an explanation of dating violence (Hotaling & Sugarman, 1986). The hypothesis is that dating violence is transmitte d from one generation to the next through exposure to violence in the family of origin during the formative years. These findings in the past have been seen as consistent with and supportive of social learning theory. Social learning theory suggests that o ne mechanism by which such behaviors may be learned is through observation and imitation of others and vicarious differential reinforcement. A childs early experience and exposure to models in the family may be a source of learning about gender role expec tations, perceptions, and attitudes toward violence. Social learning theory therefore has a potentially significant contribution to make toward an explanation of dating violence (Jackson, 1999). The intergenerational transmission model and evidence of effe cts from childhood exposure to violence on experience with violence as a young adult are consistent with general social learning theory but cannot be considered a full test of the theory ( Akers, 1998). This is true because imitation is only one element (an d typically not the strongest part) of the social learning process, which also includes differential association, definitions (attitudes), differential reinforcement, discriminative stimuli, and other elements. Also, recent and current learning experiences would be expected to have greater impact than prior events that may have occurred many years ago. However the current dissertation studies only the learning processes in the family relationship that was completed in childhood. It is also limited to measur ing experience of violence within the family.

PAGE 37

37 Social Bonding (Attachment) Theory Bowlby (1973) suggested that children form mental representations or prototypes of relationships based on their experiences with primary caregivers during childhood. These rel ationship prototypes are relatively consistent over time, and serve as templates for future relationships. He proposed what he calls attachment theory to suggest that adolescents tend to select dating partners based on these prototypes. According to atta chment theory, children develop internal working models (IWMs) of their relationship from childhood experience mainly from attachment to parents, and their attachments play a crucial role in later psychological functioning (Bowlby, 1973; Sternberg, Lamb, G uterman, Abbott & Dawud -Noursi, 2005). That is, the quality of the childparent attachment relationship powerfully affects later social development (Bowlby, 1973; Cassidy, 1994; Sroufe, Clarson, Levy & Egeland, 1999). Individuals with negative relationship templates may gravitate toward violent dating relationships. The expectation is that those with insecure attachment styles resulting from childhood mistreatment would be at particularly high risk for dating violence (Cicchetti & Howes, 1991; Critenden & Ainsworth, 1989; Schwatz, Hage, Bush & Burns, 2006). Although the two are different, this attachment theory of Bowlbys, has obvious similarity to Hirschis (1969) social bonding theory in which attac hment to others is a key component of bonds for society hypothesized to control deviance. Hirschi s proposed that the main reason of criminal behavior is weak social bond. A ttachment, commitment involvement and belief are key elements of social bond. H owe ver, this study only allows to test the attachment among Hirschi s four concept of social bonding theory. Attachment in Hirschis theory refers to close, warm and affectionate ties to parents, peers and others. Parental attachment is especially important in the theory and most relevant to this dissertation project because it is related to the respondents relationships in their families of origin. Attachment to

PAGE 38

38 parents is measured in the dissertation by current attachment. It is likely, however, that thei r current parental attachment was first formed or strongly influenced by attachment to parents in childhood. L ittle use has been made of social bonding concepts in research on dating violence. Cleveland et al. (2003) reported that females who were attached to their mothers were less likely to experience dating violence victimization. Similarly, high school students who reported higher attachment to their parents were less likely to perpetrate dating violence (Chapple & Hope, 2003). Despite a growing number of studies examining the early childhood experiences of dating violence, the relationship between childhood sexual, physical, and emotional victimization and neglect, as well as disturbances in the attachment relationship, and later perpetration and vic timization by dating violence remains unclear. This dissertation analyzes data from samples of university students in the United States and Korea for both cross -cultural and gender comparisons, both in witnessing of inter -parental violence and in experien cing child abuse in the family. It also examines the applicability of three general theories of crime and deviance (self -control, intergenerational transmission /social learning, and social bonding/attachment theory) to dating violence directly.

PAGE 39

39 CHAPTER 3 METHODOLOGY Data C ollection P rocedure Data for this study comes from research conducted by Dr. Angela Gover and myself in 2005. T he survey, entitle d Family and relationship experiences and attitudes among college students was created and administered to students at one university in the United States a nd one university in Korea The survey consisted of 167 questions on topics such as family of origin violence, attitudes toward women, stalking victimization and perpetratio n, fear of violence, dating relationship behaviors, sexual behaviors, substance use, parental attachment, and self -control. Students were provided with informed consent documentation, participated voluntarily, and were assured confidentiality through the w ithholding of identifying information (such as name and student identification number). Participants were provided with survey packets that contained the informed consent documentation, questionnaire, and the name of a contact person and telephone number f or those who might have questions after completing the survey. Respondents returned completed surveys in the envelopes they were provided to ensure additional confidentiality. The importance of the research was explained orally to respondents by researcher s, and students were not provided with extra course credit as an incentive for survey participation. The sampling method was convenience sampling in a classroom setting in both countries. The convenience sampling is a non-probability method with limitation s on its ability to obtain a representative sample. The large sample size should mitigate those limitations to some extent but do not remove them. However, the focus of this dissertation research is not to make generalizations about the frequency or demogr aphic distribution of dating violence in the two countries. Rather it is to examine the extent to which different theoretical perspectives apply to these particular samples

PAGE 40

40 in the two countries. Samples nationally or categorically representative are less important for that purpose. The process of administering the questionnaire was the same in the United States and Korea. U.S data collection by Gover was completed in a classroom setting between August and December of 2005 at the University of Florida. The f inal U.S. sample consisted of 1399 students taking liberal arts and sciences course such as criminology, psychology, economics, anthropology, English and political science. T he survey response rate was 98%.1 The Korean d ata which I collected is a replica tion of the study by Gover et al. (2008). The contents of the questionnaire and the number of questions are exactly the same and the data collection procedures were identical. The original survey instrument was written in English ; three translators includ ing me, who are fluent in both Korean and English, reviewed the translated Korean survey instrument. IRB approval for the Korean study was obtained on May 18, 2007 from the University of Florida because Korean universities and research institutes are not required to obtain IRB approval in Korea. T he survey was completed in a classroom setting between May and June of 2007. The sampling method used in this survey was convenience sampling, like the previous U.S. sampling method. Students at Kyonggi University in Korea were asked about their intention to participate in this survey before class started. Before the students were permitted to enter the class, the lecturers or professors in charge of the class obtained permission for each student to participate. St udents participated in this survey voluntarily There was no reward for participation nor penalty for withdrawal of the participation. 1 The r esponse rate was calculated by comparing the a tot al of number of completed surveys to the number of student present during class on the day s surveys were administered in the United States.

PAGE 41

41 Researchers selected classes from those that were included in the course schedule for the 2007 spring semester at Kyong gi University. Based on course information such as classroom size, major, and department offering the course, the classes that were selected had large numbers of students and were open to all students, regardless of their majors. These classes were selecte d so that students majoring in other fields were included. The average classroom size for courses open to all students was 150. In addition to these larger size classes, small courses having less than 50 students were selected as well. These courses were elective or core courses of their departments. Overall, data were collected from the colleges of humanities, law, social science, business, and natural sciences, and students from the departments of foreign languages, public administration, social work, police administration, youth studies, economics, applied statistics, industrial property, accounting, mathematics, physics, chemistry, and biology participated in this survey. Similar to the Gover survey done in 2005, the Korean survey instruments had a tota l of 167 questions with topics such as basic social demographic factors, campus life, family of origin violence, perpetration, and victimization by stalking, attitudes toward women, fear of violence on campus, perpetration, and victimization by dating viol ence, religiosity, sexual behavior, substance use and drunk driving, attachment to parents, and self -control. Before distributing the survey, the purpose and importance of the survey was explained orally by the principal investigator in front of the students. Due to the sensitivity of some of the questions, participants were provided with survey packets that contained the informed consent documentation, the questionnaire, and the name of a contact person and t elephone number for those who might have questio ns after completing the survey. Respondents returned completed surveys in the envelopes in which they were provided to ensure additional confidentiality. Although the included

PAGE 42

42 informed consent documentation explained the participants right to participate and their right to withdraw from the survey and included contact information regarding this research, all participants were offered the opportunity to have their rights and the contact information explained orally by a principal investigator before they started answering any questions. A total of 1 399 questionnaires were completed in the Korean survey. T he response rate was 96.5% .2 Measur ement of Variables Dependent Variables: Perpetration and Victimization in Dating Relationships The dependent variables in this study are: (1) perpetration of physical dating violence, (2) perpetration of psychological abuse/violence in dating relationships, (3) victimization by physical dating violence and (4) victimization by psychological abuse/vi olence in dating relationships. These were measured by items based on but modified from the Revised Conflict Tactics Scale, which is the most commonly used scale of intimate partner violence (Straus, Hamby, Boney -McCoy, & Sugarman, 1996). Questions related to perpetration of and victimization by sexual abuse in dating relationships were also asked on the instrument, but not included in the statistic analysis for this dissertation. Respondents were asked how often those behaviors happened in the past year. T hey are ordinal variables and the answers were cod ed 0 to 4 (0=never happened, 1= once, 2=twice, 3=35 times, 4=6 or more times). All items measuring the dependent variables were added to create the score of dating violence: 0 means no experience and a high er score means more experience with dating violence. Each dependent variable was measured with selected items using factor analysis. Therefore the U.S. sample and the Korean sample have a possibility that they will have different number of items to measure the same dependent variable. 2 T he response rate was calculated by comparing a total of number of completed surveys to the number of distributed surveys in the classrooms in Korea.

PAGE 43

43 Perpetration of physical violence in Korea was measured by a 9 item scale (see Appendix A odd numbered items F1-F17), with scores of 0 to 32 (alpha = .857). Perpetration of psychological abuse was measured by 6 items. It in cluded items on actions taken by the respondent against a dating partner: I threatened to hit or throw something, I prevented my partner from seeing family or friend, I insulted or swore at my partner I accused my partner of being a lousy lover, I called my partner bad names and I shouted or yelled at my partner. (refer to F19, F21, F27, F29, F31, F33 in the Appendix A ). The scale scores ranged from 0 to 24, with a moderate alpha coefficient of .652. Victimization by physical violence was m easured by 9 items. The items included descriptions of the partners behavior such as my partner threw something that could hurt, twisted my arm or hair, kicked me, slapped me, pushed or shoved me, punched or hit me with hand or an object, cho ked me, slammed me against a wall and grabbed me (refer to even number items from F2 to F18 in the Appendix A ). Reliability of selected item is .878. The scores of victimization by physical violence ranged from 0 to 36. Victimization by psychological abuse was measured using selected 6 items by factor analysis. The items included partners behaviors such as my partner prevented me from seeing family or friend, insisted on knowing my whereabouts all the time, insulted or swore at me, insisted on knowing who I was talking to on the phone, insulted or swore at me, accused me of being a lousy lover and shouted or yelled at me (refer to F22, F24, F26, F28, F30, F34 in the Appendix A ). Reliability of selected item is .762. The scores of victimiz ation by psychological abuse ranged from 0 to 20 Perpetration of physical violence in the U.S. sample was measured by 9 items (odd numbered items F1-F17). Reliability of 9 questions is .864. All 8 questions were added up to

PAGE 44

44 create a scale of perpetration of physical violence. The score is 0 to32 Perpetration of psychological abuse was measured by 6 items. It included these behaviors that the respondent reported against their partner I insulted or swore at my partner, I accused my partner of being a lou sy lover, I called my partner bad names and I shouted or yelled at my partner. ( r efer to F19, F25, F27, F29, F31, F33 in the Appendix). The alpha value is high (alpha=.818). The score of perpetration of psychological abuse ranged from 0 to 24. Victim ization by physical violence was measured by 9 items. The items included descriptions of the partners behavior such as my partner threw something that could hurt, twisted my arm or hair, kicked me, slapped me, pushed or shoved me, punched or hit me with hand or an object, choked me and slammed me against a wall (refer to even numbered items from F2 to F18 in the Appendix A ). Reliability of selected items is .845. The score of victimization by physical ranged from 0 to 36. Victimization by psychological abuse was measured using 6 items. The items included partners behaviors such as my partner prevented me from seeing family or friend, insisted on knowing me whereabouts all the time, insulted or swore at me, insisted on knowing who I was talking to on the phone, insulted or swore at me, my partner called me bad names and shouted or yelled at me (refer to F22, F24, F26, F28, F30, F32 in the Appendix A ). Reliability of selected items is .800. The score of victimization by psychol ogical abuse ranged from 0 to 2 4 All of the dependent variables were measured also for a full model including both U.S. and Korea samples. Items to measure each dependent variable were selected using factor analysis. Finally, nine items for perpetration o f physical violence (Appendix A: F1 17 odd numbers), seven items for perpetration of psychological abuse (Appendix A: F19, F21, F25, F27, F29, F31, F33), nine items for victimization by physical violence (Appendix A: F2-F18 even numbers) and

PAGE 45

45 seven items fo r victimization by psychological abuse (Appendix A: F20, F22, F26, F28, F30, F32, F34) were selected to measure dependent variables of the full model. Independent Variables Experience of violence in the family in childhood It is important to measure exp osure to violence during childhood, both witnessing violence between parents and experiencing violence as a victim, because research on the intergenerational transmission of violence has shown that there may be differential effects of witnessing violence versus directly experiencing violence (Stith et al., 2000). The measure of child abuse was created from 12 items of the Revised Conflict Tactics Scale (Straus et al., 1996) to indicate whether a respondent had ever experienced any abusive incidents at the hands of a parent, guardian, or other caretaker. Students were asked to respond with had not experienced or had experienced childhood abuse by parents, stepparents or guardians on actions such as: throw something at you that could hurt you, push, grab o r shove, pull your hair, slap or hit you, kick or bite, choke or attempt to drown you, hit with some object, punish you with a belt, board, cord, or hit you so hart that it left bruises or marks. A lthough the 12 items for child abuse have been shown to ma ke up a reliable and valid scale in most research there is a problem of using a scale without checking to see if it is reliable for the study sample. I conducted factor analysis for each sample and found the alpha values to be low in both samples. Therefore, a s the result, d irect experience of violence by Korean students was measured using 9 items selected by factor analysis among 12 Revised CTS items that were found to found to have acceptable alpha levels (refer to Appendix A:b1, b2, b3, b4, b5, b6, b7, b9,b10). Reliability of selected item for child abuse in Korea is high (alpha=.812). The score ranged from 0 to 9 because responses for each question were coded dichotomously. U.S. students direct experience with child abuse was measured by 7 items selected by factor analysis

PAGE 46

46 (refer to Appendix A: b1, b2, b3, b4, b7, b9, b10). Reliability of the selected item for child abuse in the U.S. is very acceptable (alpha=.787). The scores of direct experience of child abuse in the U.S. ranged from 0 to 8 bec ause responses to the questions were coded as dichotomous responses. An 8 item scale of child abuse was recreated for the full model with both the Korean and American samples with an alpha value of .768 (Appendix A: b 1, b2, b3, b4, b5, b8, b9, b10). T he s core of direct experience of child abuse ranged from 0 to 7. Indirect experience of violence was measured by fathers violence against mother, and mothers violence against father. Respondents were asked whether they had ever witnessed the father hit the m other or whether they had ever witnessed the mother hit the father in childhood (see Appendix A: B13, B14). Each question was dichotomous variable coded 0 and 1 (0=no, 1=yes) Th e indirect experience of violence was measured with the same questions for the Korean, U.S. and full model. Self -control Self -control was measured by a 23 item scale identical to that utilized by Grasmick et al. (1993)3. This reflects the unidimensionality of self -control proposed by Gottfredson and Hirschi (1990). As a result of factor analysis for 23 items, a lpha value of self -control for Korean sample is .871, and for American sample is .902. The items measured included impulsivity, preference for simple tasks, risk -seeking, physical activities, self -centeredness, and temper (se e items H1 to H23 in Appendix). Respondents were presented with items on each of these dimensions (e.g., I often act on the spur of the moment without stopping to think, Sometimes I will take a risk just for the fun of it) and asked to respond with one of the following: strongly agree, agree, disagree 3 T his 23 items of self control could not reflect Hirschi s (2004) re conceptualization of self control. In addition, some studies, for example (Piqueor & Bouffa rd, 2007) have included risky sexual behavior as a measurement of self control, this study used only Grasmick et al. s (1993) self control measurement R isky sexual behavior were used as a control variable separately in this study.

PAGE 47

47 or strongly disagree (coded from 1 to 4). The scores were added together to create a scale of self control. Lower scores indicated lower self -control. T he analysis also tested the separate effects of each of the six dimensions of self -control as was done by Kerley et al. (2008). Using factor analysis and reliability test, all six dimensions have each scale4. H owever, the test result of six dimensions will not be report ed mainly in this study since only tem per dimension showed the significance5. Attachment to p arents Attachment to others as a social bond was measured using two categories: attachment to mothers and attachment to fathers. To measure attachment to mothers and fathers, respondents were presente d with 4 items (see items I1 to I 4 for attachment to mothers, I5 to I8 for attachment to fathers in Appendix A ). The i tems include How often do you trust your mom/female guardian?, How often do you feel you can talk to her about your problem?, How often do you think she is genuinely interested in you?, and How often do you feel that she supports you? Response cate gories offered were never, sometimes, half of the time, usually, and always 4 Alpha value for Ko rean sample is: impulsivity =.716, simple tasks=.788, risktaking =.803, physical activities=.719, self centered=.782, temper=.744, alpha value for American sample is: impulsivity=.760, simple tasks=. 803, risk taking =.788, physical activities=.793, self c entered=.806, temper=.817 5 T he analysis also tested the separate effects of each of the six dimensions of self control as was done by Kerley et al. (2008). Using factor analysis and reliability test, all six dimensions each scale5. Impulsivity was measur ed using 4 items. Examples of questions are I often act on the spur of the moment without stopping to think, I dont devote much thought and effort to preparing for the future (for impulsivity, see Appendix A: H1 to H4). Simple task preference was meas ured using 4 items such as I frequently try to avoid projects that I know will be difficult,when things get complicated, I tend to quit or withdraw (see Appendix A: H5 to H8). Riskseeking was measured using 4 items such as I like to test my self ever y not and then by doing something a little risky, excitement and adventure are more important to me than security (for risk seeking, see Appendix A: H9 to H12). Preference for physical activities was measured using 3 items such as if I had a choice, I would almost always rather do something physical than something mental, I like go out and do things more than I like to read or contemplate ideas (for physical activities, see Appendix A: H13 to H15). Self centeredness was measured using 4 items such a s I am not very sympathetic to other people when they are having problems, if things I do upset people, it is their problem not mine (for self centeredness, see Appendix A: H16 to H19). Temper was measured using 4 items such as I lose my temper pretty easily, when I have a disagreement with someone, it is usually hard for me to talk calmly about it without getting upset (for temper, see Appendix A: H20 to 23).

PAGE 48

48 (coded from 1 to 5). Finally, the scores were added together to create a scale for attachment to mother and father. Higher scores indicated higher attachment. A lpha value of attac hment to mother for Korean sample is. 863, and attachment to father for Korean sample is 902. A nd alpha value of attachment to mother for American sample is .865 and attachment to father for American sample is .776. Control Variables Gender was coded wit h males as 0 and females as 1. Family Structure was measured by asking the student respondents about the family in which they were raised lived with two biological parents, lived with single parent, lived with stepparents, and lived with others, su ch as adoptive parents or other relatives. The variables are nominal and coded from 1 to 4 (1=two biological parents, 2=single parent, 3=stepparents, and 4=others. However, these variables were coded into dummy variables in OLS regression s The reference group is two biological parents. Age response categories were 1=18; 2=19; 3=20; 4=21; 5=22 or older, but the responses were coded as a continuous variable because the age of most of college students ranged from 18 to 22. Extent of Dating Involvement was measured by asking respondents the ir current relational status. Students were asked to select the ir current dating status from the following choices: not currently dating, occasionally dating, steady/exclusively dating, and other. O ther included more serious status options such as engaged, married, and divorced and others. The reference group is not currently dating. A lpha value of Korean sample is. 752, and American sample is .779. H owever, in using the not currently dating as a ref erence gr oup some issues arose: 1) having a dating partner is a necessary condition for dating partner ; 2) it is possible that the students have had experience of dating violence in the past although they do not have current

PAGE 49

49 dating partner ; 3) it is uncl ear whether previous bad experience of dating led the students to discontinue dating someone. D ue to these problems, it may be right to ask the relationship between whether they experienced dating violence with current dating partner. Unfortunately, the su rvey instrument could n ot reflect this concern. If the net effect of the seriousness of dating relationship was strong, I ran additional OLS regression models for only those students who have a current dating relationship as a solution for this problem. H o wever, the result of analysis for this group of respondents that excluded those without current dating relationship showed no big difference Therefore, the previous results without exclusion will be reported mainly. A dditional discussion will be Chapter 4. Three questions were asked regarding risky sexual behavior, such as, age at first sexual intercourse, a total number of sexual partners, and a total number of sexual partners during the last three months. Students were asked to select their age at first sexual intercourse among choices of no sexual intercourse; 18 and more, 17, 16, 15, 14, and 13 or younger, and this question was also coded from 0 to 6. Also, students were asked the number of sexual partner s so far and during last three months as separate questions. Students selected their answers from among no experience, 1 partner, 2 partners, 3 partners, 4 partners, 5 partners and 6 or more. That is, all three questions were rated on a scale from 0 to 6 a nd the three questions were added to create a scale of risky sexual behavior. The s core of risky sexual behavior range d from 0 to 18. A higher score means more risky sexual behavior. T his scale was applied to the Korean, U.S. and full model equally. Care s hould be taken in using the variable of risky sexual behavior as a control variable because it is possible to present a tautological issue in explaining deviant behavior (dating violence) by using deviant behavior H owever, it i s less of a tautological iss ue because we

PAGE 50

50 excluded questions related to sexual behavior as dependent variables. Moreover, because p revious literature (Gover et al., 2008) has shown deviant behaviors powerful effect on dating violence, it was left as a control variable in this study Living on campus is coded 0 and living off -campus is coded 1. Although the Univers i ty of Florida provides residence halls for most of its students, Korean universities have traditionally provided a very limited number of residence halls. Unlike the U.S. student group, most students in Korea were expected to answer that they lived off -campus. Because most of Korea is composed of Asians (Koreans), the Korean sample was not tested for races impact on dating violence. Only the U.S. sample included the race/e thnicity variable on OLS regression. Although race categories were white, Hispanic, black, Asian, and other (Pacific Islander, American Indian/Alaska Native), race was used as a dummy variable in the regression models. The reference group was white. The f u ll model combining two data groups had a country variable instead of race Korea wa s coded 0 and the U.S. wa s coded 1. Data Analyses SPSS was used in t he data analysis. Descriptive statistics such as frequencies, percent ages, and means were used to provide information on the characteristics of the respondents. Also, simple correlations among independent variables and dependent variables were carried out ( see Appendix C). Multivariate regression models were run to analyze the net effects of t he independent variables on dating violence. They were run for the Korean and U.S. samples separately and were then compared across the two samples to assess which of the theoretical models showed similar or dissimilar effects in the two countries. The mod els tested we re divided into full model, Korean model, and U.S. model. E ach of the three model s was tested using the

PAGE 51

51 Logistic Regression and Ordinary Least Square (OLS) Regression model6. Because the data of the dependent variables were clustered lower sco re, logistic regression analysis also was tested for all three models. Logistic regressions were explored to find the difference between two groups that ha d 1) never experienced dating violence in the past year and 2) who ha d experienced dating violence in the past year. However the result s indicate there were no significant differen ces overall between logistic regression result s and OLS regression results. Therefore, the result s of OLS regression will be reported mainly to provide a clear explanation of net effect and R -square. Only significant result s of logistic regression will be reported in this dissertation. Tables of L o gistics regression are located in Appendix. OLS regressions examined main theoretical factor s among all students including the two g roups in each model (full, Korean, and U.S.) first. H owever, this regression included many respondents who have ever experienced the dating violence. It means that it is hard to test the significant factors that affect seriousness of perpetration and victi mization of dating violence. In addition to the initial OLS regressions for the entire samples of each model, additional OLS regressions were run for students who had experienced perpetration of or victimization by dating violence Th ese additional subgrou p regressions will lead to examin ing significant factors that affect seriousness of perpetration and victimization of dating violence considering only those who had experienced any dating violence That is, it allows testing the models to see if they can account for variation in seriousness only among those who had engaged in or been the victim of violence, excluding those who scored zero on the scales. Moreover, it allows comparison of 6 Because the data of the dependent variables were clustered toward lower scores Logistic Regression analysis also was used for all three models. Logistic regressions were explored to find the difference between two groups : those respondents that had never experienced dating violence in the past year and those who did experience da ting violence in the past year. There were no significant differen ces overall between Logistic regression results and OLS regression results. However, result s of OLS regression will be reported mainly to provide a clear explanation of net effect and R squa re. Only significant differen ces will be reported because there were no noteworthy difference s from OLS regression.

PAGE 52

52 differences or similarities between overall models that includes those who scored zero on the scales and the group that scored higher on the scales In addition, both samples will have the full, female, and male models to study gender effects using OLS regression once gender effect is found on the entire s ample in each country. The analysis should show which set of theoretical variables apply in both societies, as well as for males and females, and which have the strongest effects7. In addition to regression models, comparisons of regression coefficients will be tested for gender and country difference. Although the coefficient values obtained in each regression model allow checking the significant net relationship among independent variables, it is hard to compare the power of independent variables using bi across models. Paternoster, Brame, Mazerolle and Piquero (1998) provided the formula for a correct statistical test of effects in two samples. Using their formula (see below), I will explore the difference between two regression coefficients obtained fo r both male and female models and for U.S. and Korea models. The formula recommended by Paternoster et al. (1998) is as follows: Z =b1 b2 SE b1 2+ SE b2 2 (3 1 ) This will allow me to compare whether independently obtained coefficients are significantly different, both between female students versus male students, and between U.S. students versus Korean students. For instance, if the obtained coefficients of self -control effects on data violence in the Korean model is 1.54 with signi ficant p value, and the U.S. model has 7 Before testing OLS regression, three assumptions such as linearity, normality and independence of linear regression were tested. A lthoug h the data violated the normality, data transformation using natural log did not affect the result of regression model compared to using the original data without transformation T herefore, the overall OLS regression will report the OLS regression models w ith original scale s without natural log transformation.

PAGE 53

53 1.79 for coefficient with significant p-value, the recommended formula will be used to decide whether the two coefficient values are significantly different or not.

PAGE 54

54 CHAPTER 4 RESULTS Korean Dating Violence Descriptive Statistics for the Korean Sample Table 4 1 presents the descriptive statistics for the Korean sample. The sample size of collected data is 1,399 students from Kyonggi University: 632 male (45.2%) and 767 female (54.8%) students participated in this survey. Among the participants, 202 students (14.4%) were 18 years of age, 322 students (23%) were 19, 245 students (17.5%) were 20, and 630 students (45%) were 21 years of age or older. Unlike students at the University of Florida, m ost students in Korea live off -campus as the university residence halls are usually either very small or not provided at all. Only 70 students (5%) in Kyonggi University lived in on-campus residence halls, almost 1 327 (95%) students lived off -campus. Near ly all students in the sample (1306, or 94.4%) answered that they were raised by two biological parents, 58 students (4.1%) stated that they were raised by single parents, 16 students (1.1%) were raised by stepparents and 19 students (1.4%) were raised by adoptive parents, other relatives and others. A total of 1 3 99 students responded to the survey question regarding current dating relationship status. Over half (50.7%) did not have a current dating partner at the time. The next largest group, 432 students (30.9%) answered they had a steady or exclusive dating partner. Roughly half that number, 211 students (15.1%) reported they were dating a partner occasionally. The balance (3.4 %) answered that they were engaged, married or divorced, or had other forms o f dating. These 47 students were merged into other in the statistical analysis. Three questions addressed risky sexual behavior such as age upon first sexual intercourse, a total of number of sexual partners and a total of number of sexual partners dur ing the last three months. A majority (61.8 %) of students answered they had never had sexual intercourse in

PAGE 55

55 response to the first question asking age upon first sexual intercourse, and nearly a third (28.1%) reported having had sexual intercourse at 18 ye ars of age or older. 3 % had experienced sexual intercourse at the age of 17 and less than 5 % reported their first sexual intercourse took place when younger than 16. In response to the question regarding a total of number of sexual partners, almost 6 1 % ( 849) of students answered they had never had sexual intercourse so far. Among 503 sexually active students, 173 reported only one sexual partner, 76 reported having two partners, 78 reported three partners, 51 reported four partners, 33 reported five partn ers and 92 reported six or more partners. Consistently, in response to the question of a total of number of sexual partners during the last three months approximately two thirds (61.3%) answered they had never had sexual intercourse so far. 11 % answered they had not had sexual intercourse during the past three months. Among students reporting sexual activity in the past three months, 276 (19.7% of the total of sample) reported one partner, 28 students had two partners, and 26 students had three partners. Child abuse was measured using 9 items, the scale of child abuse ranged from 0 to 9. Less than half of the survey group (610 students ), had 0 score, meaning they had never experienced any abusive behavior by their parents in childhood, and the mean score was 1.52. In questions regarding parents violence observed during childhood, 315 students (22.5%) answered they had witnessed their fathers violence against their mother, and 118 students (8.4%) answered they had witnessed their mothers violence against the father. Self -control was measured by a total of 23 items used by Grasmick et al. (1993). Students responded to a four answer scale (1=strongly agree; 4=strongly disagree) and a low score indicated low self -control. Self -control scores ranged from 23 to 115, and the mean score of self control is 61.67. Attachment to parents was divided into two measures: attachment to mother

PAGE 56

56 and attachment to father. The participants responded to four questions on a 5 -point scale: (1) how often do you trust your moth er/female guardian? (1=never; 5=always); the internal consistency of the scale was acceptable (alpha of attachment to mother = .863; alpha of attachment to father = .854). Scores of attachment to mother ranged from 4 to 24, and the mean score of attachmen t to mother is 16.63. Scores of attachment to father ranged from 4 to 24, and the mean score of attachment to father is 15.19. The survey addressed four major forms of dating violence: perpetration of physical violence, perpetration of psychological abuse victimization by physical violence and victimization by psychological abuse. Korean students were not frequently exposed to dating violence, yet the numbers who were involved in either perpetration of or victimization by dating violence are not insignificant. Perpetration of dating violence was measured using two forms. Perpetration of physical violence measures ranged from 0 to 32. The mean score of perpetration of physical violence was .63. The large majority, 1, 153 students (85.4%) had not perpetrated physical violence against their partner in the past year, yet 197 students (14.5%) did perpetrate physical violence against their dating partner in the past year. Perpetration of psychological abuse measures ranged from 0 to 24 and the mean score of psychological abuse was 1.08. Roughly two thirds (64.9 %) of students had never perpetrated psychological abuse against their partner in the past year, and 416 students (29.7%) did perpetrate psychological abuse against their partner in the past year. Victimiza tion by dating violence was also measured using two forms. Victimization by physical violence measures ranged from 0 to 36, and the mean score was .57. A not insignificant number of students (11.7%) reported experiencing physical violence by their partner in the past year. Victimization by psychological abuse measures ranged from 0 to 20, and the mean score of victimization by psychological abuse was 1.80. About a third (31.1 %)

PAGE 57

57 of students were victimized by their dating partner in the past year. Consiste ntly, scores of perpetration of and victimization by psychological abuse were higher than perpetration of and victimization by physical violence. Results of Multivariate Analyses for the Korean Sample The first step was to test the net effect of independen t variables of the Korean sample. OLS regression analysis was run to examine the net effect of independent variables and covariates. T hree social learning variables : direct experience of child abuse, witness of father s violence or witness of mother s violence, and low self -control, and two attachment variables were used as independent variables. Gender, age, family structure current dating relationship, living on or off campus and risky sexual behavior were used as control variables The result s of perpetration of physical violence and perpetration of psychological abuse are explained separately Perpetration of physical and psychological dating violence Table 4 2 shows the result of OLS regression of physical violence for the Korean sample. 1 304 students were used for OLS regression and nonperpetrators and perpetrators were not investigated as separate groups Overall, social learning theory and self -control theory were partly able to show predictive power regarding physical dating violence. Among three variables of social learning theory, only direct experience of child abuse showed significant influence on perpetration of physical violence. Those students who had ever been exposed to child abuse in childhood were more likely to be perpetrat ors than were other students (b=.135, p = .00). Low self -control was also significantly related to the higher likelihood of perpetration of physical violence (b= -.037, p = .00). All attachment variables failed to show a significant effect on physical violence. Age is another

PAGE 58

58 significant factor on physical violence1. Younger students were more likely to become involved in perpetration of physical violence than were older students (b= .133, p=.038) Gender did not significantly account for the perpetration of physical violence Although family structure failed to show a significant effect statistically under the criteria of p -value .05, being raised by other adoptive parents or relatives showed significant effect under .10. Students who were raised by adoptive par ents or relatives were more likely to perpetrate physical violence against their dating partner compared to students who were raised by two biological parents (b=1.118, p=.054). Students engaged in risky sexual behavior were also more likely to be perpetr ators than were students with lower scores in that area. As expected, having a dating partner increased the likelihood of perpetration of physical violence compared to students with no dating partner. Specifically s tudents having an occasional dating part ner were more likely to become perpetrators of physical violence than students having no dating partner, and having a steady dating partner also was a significant predictor for the perpetration of physical violence. However, as mentioned earlier obviously one has to be dating to have any dating violence; it is tautological to say that people who date have a greater chance of dating violence. Due to this tautological issue, an additional regression model was tested for students who reported a current dating relationship. Among 1 304 students, only 641 students were analyzed for the additional regression model to test net effect of dating seriousness on physical violence. Other variables used were identical to those used in the previous OLS regression model The result s of the subsample of students in a serious dating relationship, defined as engaged, married, or divorced, show ed a significant positive effect on physical violence compared to those who reported having 1 Age failed to show a significant impact in the logistic regression models

PAGE 59

59 an occasional dating partner (b=1.38, p=.01 ). However, there was no difference between students having a steady or exclusive dating partner and students having an occasional dating partner. The OLS regression model of perpetration of psychological abuse (Table 4 -2 ) shows similar net effects of inde pendent variables with respect to physical violence. A total of 1,278 samples were used for this analysis Only the direct experience of child abuse can offer a partial explanation for the perpetration of psychological abuse (b=.163, p = .00) Witnessing a f athers violence against mother, or a mothers violence against father, as well as measures of attachment to mother, or attachment to father, all failed to show predictive powers in the identification of perpetration of psychological abuse. Low self -control is very significantly related to psychological abuse (b=.041, p = .00) As with the previous physical violence model, having a dating partner of any type increased the likelihood of being a perpetrator of psychological abuse, and risky sexual behavior play s a very significant role in predicting the perpetration of psychological abuse. Gender difference was found in measures of perpetration of psychological abuse. Females were more likely to perpetrate psychological abuse against their partners (b=.361, p=.0 06) Also, interestingly, age has a reverse effect on perpetration of psychological abuse compared to perpetration of physical violence. Unlike perpetration of physical abuse which showed younger students were more likely to be a perpetrator, it was found that older students are more involved in perpetration of psychological abuse against their partner. Overall, R2 is .151 and the model fits well. Risky sexual behavior also sh owed a significant effect on perpetration of psychological abuse ( b= .171, p=.00) As expected, having a dating partner increased the likelihood of perpetration of psychological abuse. Due to the tautological issue s the subsample of students having a datin g partner was extracted to test the effect of dating seriousness on perpetration of psychological abuse. A tot al of 633 students were included in additional OLS

PAGE 60

60 regression. Results failed to show a significant effect of the seriousness of a dating relation ship on perpetration of psychological abuse. In sum, perpetration among Korean students is significantly explained by direct experience of child abuse, self -control, age and having risky sexual behavior D irect experience of child abuse, and risky sexual behavior are positively related to perpetration of physical violence and psychological abuse. Low self -control showed negative relationship s with perpetration of physical violence and psychological abuse. Although the effect s of age are found in both form of perpetration, the direct ion is reversed. Y ounger students w ere more likely to perpetrate physical violence against their dating partners, and older students w ere more likely to commit psychological abuse against their part ners Gender effect s are found only in psychological abuse. Although both perpetration forms show the significance of dating relationships the effect of being involved in a serious dating relationship (engaged, married or divorced) decreased or disappeare d for the subsample which selected only students having a dating partner currently. A ttachment to parents and living on or off campus failed to show significance on perpetration of dating violence in the Korean sample. Physical and psychological victimiz ation by dating partners A tot al sample of 1,308 respondents was used in the regression model of victimization by physical violence. Although the model fits well, the R2 is relatively low (R2=.056). On a limited basis, social learning theory and measures o f self -control show significant predictive power regarding victimization by physical violence (see Table 4 3 ). Attachment to mother, attachment to father, and witness of parents violence against their partners failed to show predictive power with respect to victimization by physical violence. Those students who had experienced child abuse by their parents were more likely to be victimized by physical violence than others (b=.104, p=.005). T hose students who had lower scores on self -control were more likely to be

PAGE 61

61 victimized by physical violence than others (b= -.033, p=.000) Nor was there found any effect of gender2, age, family structure, and risky sexual behavior. However, living on-campus increased the likelihood of being a victim by physical violence ( b= .787, p=.014). Although having a steady or exclusive dating partner bore a significant relationship to being a victim by physical violence, having an occasional dating partner revealed no effect on victimization by physical violence Due to the tautologic al issue inherent in compar isons between students without a dating partner and students having a dating partner, the seriousness of the dating relationship was tested additionally after selecting students who ha d a current dating partner. Ne ither having a steady or exclusive dating partner n or having more serious attachments such as engaged, married, or divorced showed a significant effect on victimization by physical violence compared to students having only an occasional dating partner. Table 4 3 also pr esents the results of the OLS regression model in predicting victimization by psychological abuse. 1,275 samples were used for this analysis. Direct experience of child abuse by parents and low self -control scores consistently showed predictive power on vi ctimization by psychological abuse. More students abused by parents in childhood were more likely to be victimized ( b=.199, p=.00). S tudents having lower self -control were more likely to be victims of psychological abuse (b=.056, p=.00) Conversely, indirect experience of family violence, such as witnessing the fathers or mothers violence, and insecure attachment to parents did not show a significant effect to predict victimization by psychological abuse. Although family structure cann ot show a net effect under critical area .05 of p -value, being raised by single parents showed a significant effect under critical area 10 of pvalue. Students who were raised 2A lthough OLS regression of victimization by physical violence failed to find a gender effect, logistic regression of victimization by physical violence in Korea found a significant gender effect. Being female increased the odds of being a victim by 71%.

PAGE 62

62 by single parents were more likely to be victimized by psychological dating violence than students who were raised by two biological parent s ( b=.884, p=.062). Other covariates show very significant net effect on victimization by psychological abuse. Gender, age, involvement in dating, living off -campus, and risky sexual behavior are very useful predictive factors of victimization by psychological abuse. Females were more likely to report higher victimization by psychological abuse (b=.820, p=.00)3. Similar to the previous perpetration model of psychological abuse, older students showe d higher victimization by psychological abuse (b=.265, p=.003) Also, students living on-campus were more exposed to psychological abuse by their partners than students who lived off-campus (b= .373, p=.003) Risky sexual behavior is a very positive factor of victimization by psychological abuse ( b=.221, p=.00) Having any dating partner, whether occasional or steady or exclusive increased the likelihood of being a victim by psychological abuse. Due to the tautological issue inherent in compari sons between students without a dating partner and students having a dating partner, the seriousness of the dating relationship was tested additionally after selecting students who ha d a current dating partner A total of 627 students had a current dating partner in the Korean sample, the seriousness of the relationship failed to show a significant effect Ne ither having a steady or exclusive dating partner n or having more serious attachments such as engaged, married, or divorced showed a significant effect on victimi zation by physical violence compared to students having only an occasional dating partner. In sum, among three variables of social learning theory, direct experience of child abuse showed a significant effect on victimization by dating violence in Korea Low s elf -control wa s also very significant to predict both victimizations by physical violence and psychological abuse. 3 In Logistic Regression of victimization by psychological abuse being female increased the odds of victimization by psychological abuse by 65%.

PAGE 63

63 Although age, sex, and being raised by single parents did not show the significance on victimization by physical violence, all variabl es were significant in the victimization by psychological abuse. The effect of seriousness of dating relationship disappeared after exclud ing students who did not have a dating partner. The results have reported the net effects of independent variables on perpetration and victimization by dating violence among Korean students regardless to exposure of dating violence in the past year. P erpetration of physical and psychological dating violence : the subgroup of perpetrators only T he result of next analysis shows the effects independent variables only among subsample of students who had experience dating violence, excluding students in score 0 on dating violence. Further regression analyses were run to test the net effect of dating violen ce for students who had ever been exposed to dating violence. Because there are two forms of physical violence and psychological abuse to measure the dating violence in this dissertation each group of perpetrator s of physical violence and perpetrator s of psychological abuse was selected separately Using OLS regression, this analysis found the significant factors to predict the perpetration among perpetrators of dating violence. Table 4 4 shows the OLS regression result selected only for perpetrators of physical violence. Among the entire sample of 1,399, only 193 students perpetrated physical violence against their partner in the past year. T his is the model of perpetration of physical violence among physical perpetrators. R2 is .174 and model is very s ignificantly fit. S ocial learning variables, including direct child abuse experience, witness of parents violence, attachment to parent, gender, family structures in which students were raised, and risky sexual behavior were not related to perpetration of physical violence. Among those students who

PAGE 64

64 had committed physical violence against their partners, low self -control, age, and serious dating involvement were revealed as significant factors of level of dating violence That is, students with low self -co ntrol were more likely perpetrators of dating violence ( b= .108, p=.015) Younger students were more frequently (seriously) to be perpetrators of physical violence than older students ( b= .727, p=.032) Also, reporting the other dating format such as engaged, married, or divorced, was a very significant factor in predicting the perpetration of physical violence. This analysis also has a tautological issue of the variable of dating involvement Due to the tautological issue inherent in compari sons between students without a dating partner and students having a dating partner, the seriousness of the dating relationship was tested additionally after selecting students who ha d a current dating partner. A total of 140 students had a current dating partner in th e Korean subsample of physical violence perpetrators, the seriousness of the relationship failed to show a significant effect Those students having a serious dating involvement such as engaged, married, or divorced were more likely to perpetrate physical violence against their dating partner compared to students having only a n occasional dating partner. H owever, there was no difference between the group having a steady or exclusive dating partner and the group having an occasional dating partner4. P erpet ration of psychological abuse was also tested. Table 4 4 shows the results of perpetration of psychological abuse by a psychological abuser. Among the entire sample, 400 students reported perpetration of psychological abuse against their partner in the past year. T his is the model of perpetration of psychological violence among psychological perpetrators. R2 is .169 and model fits well. In cases of perpetration of psychological abuse, more variables 4 In addition, age effect disappeared after selected students who have a dating partner currently. O nly self control and having serious dating involvement are significant under .05. H owev er, the effect of direct experience of child abuse (b=.420, p=.075) and family structure (being raised by adoptive parents or relatives) (b=6.57, p=.071) increased the level of statistical significance

PAGE 65

65 contribute to predict the perpetration. In the perpetr ator group of psychological abusers, direct experience of child abuse, low self -control, single parent upbringing, occasional dating, and other dating status (such as engaged, married, or divorced), and living off -campus showed significant net effects. I ndirect experience of violence, attachment, gender, age, risky sexual behavior all failed to show significance If students reported more experience of child abuse by parents in childhood, they were more likely to be perpetrators ( b=.199, p=.006) Low sel f-control significantly increased the likelihood of perpetration of psychological abuse ( b= .079, p=.000) Students who were raised by single parents were more exposed to perpetration of psychological abuse compared to students who were raised by two biol ogical parents ( b=1.770, p=.009) However, other family structures were not significant. Also, students living off -campus were more likely to be perpetrators of psychological abuse ( b= 1.161, p=.022) In terms of dating involvement, having an occasional da ting partner or other in volvement such as engaged, married, or divorced increased the perpetration of psychological abuse compared to having no dating partner. No difference was found between having a steady or exclusive dating partner and not having a da ting partner. Due to the tautological issue inherent in compari sons between students without a dating partner and students having a dating partner, the seriousness of the dating relationship was tested additionally after selecting students who ha d a curren t dating partner. A total of 271 students had a current dating partner in the Korean subsample of psychological abuse perpetrators, the seriousness of dating relationship showed a significant effect when compar ing students who ha d an occasional dating part ner to students who ha d other form s of attachment such as engaged, married, or divorced. T hat is, more serious dating relationships increased the likelihood of perpetration of psychological abuse among perpetrators of

PAGE 66

66 psychological abuse in Korea. H owever, the standardized coefficients (beta) decreased in the subsample model compared to the entire sample. Victimization by physical and psychological dating violence : the subgroup of victims only Further regression analyses were run to test the net eff ect of dating violence for students who had ever been exposed to victimization by dating violence. Because there are two forms of physical violence and psychological abuse to measure dating violence in this dissertation each group of victims of physical v iolence and victims of psychological abuse was selected separately Using OLS regression, this analysis was run to find the predictive factors of victimization among victims of dating violence. Table 4 5 shows the OLS regression result selected only for victims of physical violence. Among the entire sample, 1 58 students were victimized physically by their partner in the past year. Victimization models show a similar pattern to the previous perpetrator models. T his is the model of victimization by physica l violence among victims by physical violence. Goodness of fit has significant score, and R2 is .20. Theoretical factors failed to explain the victimization by physical violence under .05 of pvalue Only reporting other format of dating (b= 8.019, p=.001) or on-campus living (b= 5.030, p=.013) had a significant net effect on victimization by physical violence under .05. However, direct experience of child abuse ( b=.395, p=.082), low self -control (b= .089, p=.090) and gender (b= 1.737, b=.085) are moderate ly significant under .10 criteria of p -value. T he direct experience of child abuse influenced positively for physical victimization among victims Low elf -control influenced negatively for physical victimization among victims of physical violence. Being male increased the likelihood of victimization by physical violence among victims of physical violence. According to the results, reporting other dating status such as engaged, married, or divorced increased the likelihood of being a vic tim by physical violence. However it also require d additional analysis to solve tautological issues. Due

PAGE 67

67 to the tautological issue inherent in compari sons between students without a dating partner and students having a dating partner, the seriousness of t he dating relationship was tested additionally after selecting students who ha d a current dating partner. A total of 110 students had a current dating partner in the Korean subsample of physical violence v ictims. T he seriousness of the dating relationship showed significant effect when compar ing students who ha d an occasional dating partner to students who reported other relationships such as engaged, married, or divorced (b=8.629, p=.001)5. H owever, the standardized coefficients (beta) decreased in the subsample model compared to the entire sample. Table 4 5 also shows the OLS regression result selected only for victims of psychological abuse. A total of 419 samples were used in this regression model. T his is the model of victimization by psychological abuse among victims by psychological abuse R2 is .147 and goodness of fit of this model is significant (F=4.342, p=.000). Victimization by psychological abuse is explained in part by social learning theory and in part by attachment theory and self control. Those students who were abused in childhood were more likely to be victims of psychological abuse ( b=.208, p=.045) Students with low self -control were more likely to be victims compared to those who reported higher self -control ( b= -.061, p=.022) Althou gh attachment to mother failed to provide a significant explanation, an unsecure attachment to father was significantly predictive of victimization by psychological abuse among victims of psychological abuse (b= -.134, p=.024) That is, those students who f ormed an unsecure attachment to the father were more likely to be victims than were other students having a higher score on attachment to the father. In addition to these independent variables, sex, dating 5 This regression model rather increased the signific ance of direct experience of child abuse (b=.561, p=.063) and self control (b=.213, p=.008) as decreased the gender effect (b= .398, p=.742) and living on or offcampus ( b= 4.325, p=.086).

PAGE 68

68 involvement, and risky sexual behavior are signifi cant factors. Females were more likely to be victims in the victim group ( b=.992, p=.030) Students having higher scores on risky sexual behavior were more involved in victimization by psychological abuse. Having a steady or exclusive dating partner or having another dating form such as engaged, married, or divorced affect victimization by psychological abuse compared to students having no dating partner. It is a similar pattern to the previous physical violence victimization model showing that students in serious relationships were exposed to serious victimization by physical violence. Once students become involved in more serious relationships beyond occasional dating, it is hard to discontinue the relationship even when their partner is violent or abusive Due to the tautological issue inherent in compari sons between students without a dating partner and students having a dating partner, the seriousness of the dating relationship was tested additionally after selecting students who ha d a current dating pa rtner. A total of 289 students had a current dating partner in Korean subsample of psychological abuse victims E ither having a steady/exclusive dating partner or have other format such as engaged, married, or divorced failed to show a significant effect on victimization by psychological abuse compared to students having an occasional dating partner6. W e have tested the separated models for perpetrators and victims as excluding students who have not experienced dating violence. O bviously there are some similarities and differences between the entire sample and subsample s of perpetrators and victims In case s of perpetration of physical violence, low self -control, engag ing in risky sexual behavior and younger age are common significant factors between the entire sample and the subsample of physical 6 This model rather decreased the effect of direct effect of child abuse, attachment to father. T hose two variables were not significant in this model. H owever, family structure showed the significant effect on victimization by psychological abuse among psychological victims when we selected only students who have a dati ng partner. B eing raised by stepparents increased the odds of victimization by psychological abuse compared to being raised by two biological parents (b=3.414, p=.056).

PAGE 69

69 perpetrators. However, direct experiences of child abuse, risky sexual behavior were not found to have significant effect in the group of perpetrator s of physical violence. In case s of perpetration of psycholog ical abuse, gender, risky sexual behavior and age explained differently, and direct experience of child abuse, self -control, being raised by single parents explained similarly perpetration of psychological abuse. Although previous entire model reported that being female, older students engag ing in risky sexual behavior increased the likelihood of perpetration of psychological abuse all three variables failed to show the significance in the model of perpetrators of psychological abuse. Victimization mode ls also have similarities and differences between the entire student sample and the subsample of perpetrators and victims Direct experience of child abuse, low self -control, and living on -campus were significantly related to victimization by physical violence commonly in both samples. H owever, the victimization model among victims of physical violence lost the net effect of risky sexual behavior. O n the other hand, the victimization model of physical victim s showed the significant effect of gender. M ale students were more likely to be a victim than females among victims of physical violence. In case s of victimization by psychological abuse, there is an interesting finding. T he victim model among victims of psychological abuse found the effect of unsecur e attachment to father. Child abuse, self -control, being female and risky sexual behavior commonly showed the influence on victimization by psychological abuse. In sum, part of social learning theory ( especially direct experience of child abuse) and self control theory have provided significant explanations for the perpetration of and victimization by dating violence. H owever, attachment theory was supported by only victims of psychological abuse. Although all Korean models did not have gender effect on dating violence, some models

PAGE 70

70 indicated the net effect of gender group. N ext section will explore how they are different using the separated gender group. Dating Violence by Gender in Korea Gender effect was found in the perpetration of psychological abuse and victimization by psychological abuse for the entire sample, and victimization by psychological abuse for the victim group. First, T able 4 16 is showing the separated gender group for perpetration of psychological abuse in Korea. T he male perpetration m odel of psychological abuse correlated to direct experience of child abuse, attachment to mother, low self -control, having a serious dating partner, and risky sexual behavior. Interestingly, among male students, unsecure attachment to the mother was relate d to perpetration of psychological abuse. For female students, direct experience of child abuse, low self -control, having a serious dating partner and risky sexual behavior were as significant as they were to the male group. However, female students who were raised by single parents were more likely to become involved in perpetration of psychological abuse than were female students who were raised by two biological parents. Second, Table 4 16 is showing the separated gender group for victimization by psychological abuse in Korea. The result of v ictimization by psychological abuse also showed gender effect Predictive factors of male victimization by psychological abuse were direct experience of child abuse, low self -control, having a dating partner, and risky sexual behavior. Also, male students who were raised by single parents were more likely to be vict imized compared to male students who were raised by two biological parents. The female victimization model of psychological abuse had different significant factors such as self -control and age in addition to common factors such as having a serious dating partner and risky sexual behavior. Whereas the male victimization model of psychological abuse was not affected by age, the

PAGE 71

71 females age was a predictive factor. Older female students were more likely to report victimization by psychological abuse. This might be related to the dating experience. As having a serious dating partner is significantly related to victimization, older female students might be exposed to more experiences of serious dating relationships compared to younger female students. Expe rience of child abuse was not a significant factor to be related to victimization by psychological abuse. Third, T able 4 18 is showing the separated gender group for victimization by psychological abuse among victim group in Korea. T he victimization by ps ychological abuse among students who had ever been victimized by their partner showed gender difference. Male victim group members were significantly predicted by direct experience of child abuse, as well as by serious dating relationships such as being engaged, married or divorced, or engaging in risky sexual behavior. Low self -control was never a factor for male victims of psychological abuse. However, female victims were defined by low self -control and risky sexual behavior. Unlike the male victim group, neither direct experience of child abuse nor reported dating involvement was significantly related to the model of female victims of psychological abuse. I have investigated the perpetration and victimization of dating violence for Korean samples so far. I n brief, Korean dating violence, theoretically, is explained by part of social learning and self -control. In particular direct experience of child abuse and low self -control are very significant factors to explain the perpetration and victimization of dating violence. Female students are more likely to be a perpetrator and victim A nd male students are less likely to perpetrate physical violence compared to female students Younger students showed the tendency to get involved in physical violence eithe r as a perpetrator or a victim and older students showed the tendency to get involved in psychological violence. Although living on or

PAGE 72

72 off campus was a meaningful variables, it is open to question due to small number of students who stay at residency hal ls in Korea university. Now I will look at the same variables and conduct a similar statistical analysis for the American sample. Race variables will be examined in the American models. American Dating Violence Descriptive S tatistics for the American Sample Table 4 6 presents descriptive statistics for the U.S. sample. O f the total sample 1 337 in the University of Florida, 41 were coded as missing value leaving the number of usable questionnaires as 1296. There were 446 male (34.4%) and 850 female respondents (65.6%) Amo ng them, 64.7 % students were White, nonHispanic; 182 (14%) were White, Hispanic; 168 (13 %) were black; 4.8 % were Asian, less than 1 % were Pacific Islander or American Indian / Alaskan Native; and 1.8 % of students races were not known. Pacific Islander, American Indian/ Alaskan Native and other or unknown races will be shown as the others variable on the following results. In the sample group, 488 students (37.3%) lived in residence halls at the university and 808 students (62.3%) lived off -campus. The distribution by class was, 449 (34.6%) freshmen, 302 (23.3%) sophomores, 314 (24.2 %) juniors, and 228 (17.6 %) seniors. The a ge distribution was very similar to the class distribution. 18 year -olds numbered 444 (34.3 %), 25.5 % were 19 years old, 18.6 % were 20 years old and 278 (21.5 %) were 21 years old and older. A slight majority (39.5 %) of students did not have any dating partner at the time; 236 students (18.2%) reported an occasional dating partner, 490 students (37.8%) had a steady or exclusive dating partner. 58 students reported their dating status as engaged, married or divorced. The most common family structure in which the students were raised was an intact family with two biological parents: 977 students, or 75.4%, reported growing up in such a household, and

PAGE 73

73 190 students (14.7 %) answered they were raised by a single parent. A smaller group of 101 students (7.8 %) were raised by a parent and a stepparent, and 28 students (2.2 %) were raised by adoptive parent s or other relatives. Three questions were posed to measure risky sexual behavior: the subjects age at the first sexual experience, a total of number of sexual partners so far, and the total number of sexual partners during the past three months. Over one third of the respondents (464 students, or 36.6 %) did not have any previous experience of sexual intercourse, 286 students (22.6 %) had experienced sexual intercourse by 18 years of age, a smaller group (16.4%) reported sexual intercourse by 17 years, and 13.3 % of students reported sexual intercourse by 16 years age. Of the youngest group, 73 students (5.8%) identified their ages upon the first sexual experience as 15, 48 students (3.8%) were 14, and 19 students (1.5%) were 13 or younger. F or the qu estion of tot al number of sexual partner so far, c onsistently, about one third of students (461) answered they have not experienced sexual intercourse so far About one -fifth, or 285 students, have had one partner, 136 students (10.5 %) have had two partne rs, 95 students (7.3%) have had three partners, 71 students (5.5%) have had four partners, 51 students (3.9%) have had five partners, 161 students (12.4%) have had six or more partners. Within the past three months, 602 students have had sexual intercourse 529 students (40.9%) have had one sexual partner for past three months, and 79 students (6.1%) have had two sexual partners. Students who have had more than three sexual partners during the past three months are less than 5 % Child abuse was measured using seven items, the scale of child abuse measures ranged from 0 to 7. Less than half (544 students, or 42.5%) received a 0 score, meaning that they answered that they had never experienced any abusive behavior by their parents in childhood,

PAGE 74

74 and the mea n score was 1.50. Regarding questions about a parents violence, 85 students (6.6%) answered that they had watched a fathers violence against a mother, and 98 students (7.6%) stated they had witnessed a mothers violence against a father. Self -control w as measured by a total of 23 items used by Grasmick et al. (1993). Students responded to questions according to a four answer scale (1=strongly agree; 4=strongly disagree) and a low score indicated low self -control. The score of self -control measure range d from 23 to 115, and the mean score of self -control is 68.43. Attachment to parents was divided into two measures: attachment to mother and attachment to father. To measure attachment to parents, participants responded to four questions on a 5 -point sc ale: (1) how often do you trust your mother/female guardian? (1 = never; 5 = always). Internal consistency of the scale was acceptable (alpha of attachment to mother = .865; alpha of attachment to father = .776). Scores of attachment to mother ranged f rom 4 to 24, and the mean score of attachment to mother is 17.79. Scores of attachment to father ranged from 4 to 24, and the mean score of attachment to father is 16.27. Four major forms of dating violence were measured in questions in this study. The y were perpetration of physical violence, perpetration of psychological abuse, victimization by physical violence and victimization by psychological abuse. The numbers of subjects who were exposed to perpetration of or victimization by dating violence are not insignificant. Perpetration of dating violence was measured using two forms. Perpetration of physical violence measure ranged from 0 to 36. The mean score of perpetration of physical violence is 1.25. Among the 1, 192 valid samples, 895 students (75.1%) had never perpetrated violence against their partner in the past year, and 297 students (24.9%) have perpetrated physical violence by their dating partner in the past year. Perpetration of psychological abuse measures ranged from 0 to 24, and the mean sco re

PAGE 75

75 of psychological abuse 3.83. Almost half of the respondents (537 students, or 47%) had never perpetrated psychological abuse against their partners in the past year, and just over half, 605 students (53 %) have perpetrated psychological abuse against th eir partner in the past year. Victimization by dating violence was also measured using two forms. Victimization by physical violence measures ranged from 0 to 36, and the mean score is .89. Some 226 students (18.9%) experienced physical violence by thei r partner in the past year. Victimization by psychological abuse measures ranged from 0 to 24, and the mean score of victimization by psychological abuse is 3.04. U nder half (45.6 %) of students reported psychological victimization by their dating partner in the past year. Consistently, scores of perpetration and victimization by psychological abuse are higher than perpetration and victimization by physical violence. Results of Multivariate Analysis for the American Sample T he result of multivariate analys is for the American sample will be reported similarly to the result of Korean sample. The first step was to test the net effect of independent variables of the Korean sample. OLS regression analysis was run to examine the net effect of independent variables and covariates. T hree social learning variables such as direct experience of child abuse, witness of father s violence, witness of mother s violence, self -control, and two attachment variables were used as independent variables. Gender, age, family structure current dating relationship, living on or off campus and risky sexual behavior were used as control variables Additionally the race variable was inserted for the U .S. analysis L ike to Korean result of six -dimensionality of self -control, only temper was significantly related to four dependent variables in the U.S. Therefore, the result of six-dimensionality of self control is not reported in dissertation as suppor ting for unidimensionality of self -control. The result of perpetration of physical violence and perpetration of psychological abuse are explained separately

PAGE 76

76 P erpetration of physical and psychological dating violence Table 4 7 shows the result of OLS regr ession of physical violence for American sample. A total of 1 158 sample s were run to explore the net effect among variables of perpetration of physical violence, and it w as not been investigated separately for non perpetrators and perpetrators D irect e xperience of child abuse, low self -control, gender, race, dating involvement and risky sexual behavior were related to perpetration of physical violence in the U.S. T hose students who had higher scores on child abuse experiences have shown higher scores on perpetration of physical violence than lower score students, and the relationship is statistically very significant (b=.208, p=.001) Other social learning variables such as witnessing of fathers violence and mothers violence did not show any significant relationship. However, consistently, low self -control played a significant role in predicting perpetration of physical violence ( b= -.068, p=.000) Unlike previous studies that have reported that males were more likely to be a perpetrator, females were mor e likely to be a perpetrator in physical violence ( b=1.258, p=.000)7. Compared to white students, black students were more likely to become engaged in perpetration of physical violence ( b=.742, p=.029).8 This result is consistent with previous studies that reported a higher rate of perpetration of African American students (Makepeace, 1987; O Keefe, 1997). R isky sexual behavior is a very significant factor to predict perpetration of physical violence ( b= .124, p=.000) Although having an occasional dating partner did not explain the differences with those who reported no dating at all, more serious relationships such as steady or exclusive dating, and other relationships such as engaged, married, or divorc ed, were very 7 A ccording to the logistic regression result of perpetration of physical violence, being female increased the odds by 150% than being male. 8 The result of logistic regression of perpetration of physical violence failed to show a significan t race effect

PAGE 77

77 predictive for physical violence. That is, those students in serious dating relationships were more likely to perpetrate physical violence against their partner compared to other students who reported having no dating partner in the past year H owever, the problem of the tautological issue require d additional analysis that included only students who reported a current dating partner. A mong 1 158 students, 670 students were selected for additional analysis, and it allowed testing whether or not the current dating relationship wa s related to past experience of dating violence. The result indicated that all significant factors were identical to the entire sample. M oreover it showed that having a steady or exclusive dating partner increased the li kelihood of perpetration of physical violence compared to having a n occasional dating partner ( b=.688, p=.052). Other dating formats such as engaged, married, or divorced showed a significant effect on perpetration of physical violence (b=1.353, p=.049). H owever, the standardized coefficients (beta) decreased in the subsample model compared to the entire sample. This pattern was similarly found in perpetration of psychological abuse. Table 4 7 also shows the result of perpetration of psychological abuse for the U.S. sample. 1 ,111 samples were used for this model. Overall, direct experience of child abuse, self -control, gender family structure, dating involvement and risky sexual behavior showed significa nt predictability. T hose students who ha d experienced child abuse in childhood (b=.389, p=.000) w ere more likely to perpetrate their partner psychologically. S tudents with low self -control w ere more likely to be a perpetrator of psychological abuse compar ed to students with higher self -control ( b= .086, p=.000). Unlike previous research, however, including the perpetration of physical violence studied in the previous model, females were more likely to be perpetrators of psychological abuse against their d ating partners ( b=.1636, p=.000)9. Higher scores on risky sexual behavior 9 T he result of logistic regression indicated that being fem ales increased the odds of perpetration of psychological abuse by 98% compared to being male.

PAGE 78

78 are a very significant factor ( b=.382, p=.000) and having a steady / exclusive ( b=1.781, p=.000) or more serious dating relationship ( b=2.573, p=.000) also increases the likelihood o f being a perpetrator of psychological abuse. A lthough the family structure showed relatively weak relations statistically, those who were raised under stepparents w ere more likely to be a perpetrator than others who were raised under two biological parent s (b=.951, b .084). Attachment to parents, race10, and the structure of family upbringing, and living on -campus were not factors related to being a perpetrator of psychological abuse. Goodness of fit of this model is very significant and the R2 is .233. Ho wever, this model also requires additional analysis due to tautological issue of dating involvement. A dditionally 681 students who had a dating partner were selected The OLS regression indicated that having a steady/exclusive dating partner (b=1.107, p=.013) and /or having other format of dating (b=1.980, p=.018) increased the likelihood of perpetration of psychological abuse compared to having an occasional dating partner. H owever, the standardi zed coefficients (beta) decreased in subsample model compared to the one of entire sample. Although this study investigated the perpetration into two formats, the results indicate that the significant factors on perpetration is similar in the U.S. Direct experience of child abuse and low self -control showed significant power as theoretical variables, and other variables such as witness of inter -parental violence or attachment to parents failed to show the predictive power on perpetration of dating violence Also, both models reported females w ere more likely to perpetrate violence against their partner either physical ly or psychologically, and engaging in risky sexual behavior was related to perpetration of both forms of dating violence. Significant 10 Although OLS regression of perpetration of psychological abuse failed to show the race difference, logistic regression of perpetration of psychological abuse re ported that Hispanic students are less likely to perpetrate psychologically than white students

PAGE 79

79 differe nces between physical perpetration and psychological perpetration we re the factors of race and family structure. T h e model of physical perpetration indicated that black students were more likely to be a perpetrator of physical violence than white students; however, the race difference was not found at OLS regression of psychological perpetration. I n addition, being raised under step parents influenced an increase in the likelihood of perpetration of psychological abuse, but not for perpetration of physical v iolence. Now I will look at the same variables and conduct a similar statistical analysis for victimization model of dating violence in the U.S. P hysical and psychological victimization by dating partner S imilarly, victimization by dating violence is ex plained by two forms of dating violence. Physical victimization and psychological victimization asked separately and OLS regressions were run to find net effects on victimization by dating violence. The result of OLS regression showing here has not been investigated separately for non-victim s and victims T able 4 8 shows the result of OLS regression of victimization by physical violence. Overall 1 161 samples were used for OLS regressio n, goodness of fit (F=6.828, p=000) and R2 is moderate (.107). The result of victimization by physical violence reported that direct experience of child abuse, self -control, gender, race, risky sexual behavior and dating involvement are significant factor to predict the victimization by physical violence. Direct experience of child abuse showed the positive relation to victimization by physical violence, and very significant (b=.188, p=.001). Self -control showed the negative relation to victimization by physical violence. That is, students who had lower self -control w ere more likely to be a victim by physical dating violence than other students who had higher self -control (b= .041, b=.000). Although gender effect (b=.334, p=.078) i s not strongly predictable statistically still it showed the positive relationship to victimization by physical violence. Female students w ere more likely

PAGE 80

80 to be a victim of physical dating violence than male students (b=334, p=.078)11. In cases of victimization by physical abuse, members of the other race category (which includes Pacific Islander, American Indian or Alaskan Native and other) were more likely to be victimized by physical violence compared to white students (b=1.118, p= .043) Students engaged in risky sexual behavior were more likely to be victims of physical violence, and the Beta value (.156) of risky sexual behavior was the largest among the significant variables in this model. Having a serious relationship such as a steady or exclusive dating partner significantly increased the likelihood of victimization by physical violence, and other forms of relationships such as being engaged, married, or divorced sharply increased the likelihood of being a victim compared to having no dating partner. Due to the tautological issue inherent in compari sons between students without a dating partner and students having a dating partner, the seriousness of the dating relationship was tested additionally after selecting students who ha d a current dating partner. A total of 701 students had a current dating partner in the U.S. subsample of physical dating violence victims Having other format such as engaged, married, or divorced increase the likelihood of victimization by physi cal violence compared to having an occasional dating partner(b=1.396, p=.019). H owever, or having a steady/exclusive dating partner failed to show the significant effect on victimization by psychological abuse compared to students having an occasional da ting partner. Overall the standardized coefficients (beta) decreased in subsample model compared to the one of entire sample. M oreover, gender effect and race effect (other variable) that has shown in OLS regression of entire sample disappeared in this s ubsample model. 11 T he result of logistic regression failed to find gender and race difference on victimization by physical violence.

PAGE 81

81 T he psychological victimization shows similar results to physical victimization (Table 4 8 ). A total of 1,118 samples were analyzed for the model of psychological victimization in the U.S. M odel fit is good (F=13.329, p=.000), and R2 is .195. T he result of victimization by psychological victimization reports that indirect experience of child abuse, low self -control, gender, risky sexual behavior and dating involvement were related to psychological victimization Indirect violence lear ning from parents, attachment to parents, age, race, family structure, and living on or off campus did not affect the likelihood of victimization by psychological abuse. Beta value (.326) of risky sexual behavior was largest among the significant variables in this model. Students who ha d experienced more direct child abuse in childhood were more likely to be a vic tim of psychological dating violence (b=.296, p=.000). Lower self -control increased the likelihood of being a victim by psychological abuse (b = .056, p =.000). Gender effect wa s also significant Female students were more likely to be victimized than male students psychologically by their dating partner (b=.720, p=.011)12. Engaging in risky sexual behaviors (b=.383, p=.000 ) significantly predicted victimization by psychological abuse. Also, dating involvement is a significant factor to predict victimization by psychological abuse. However repeatedly d ue to the tautological issue inherent in compari sons between students without a dating partner and students having a dating partner, the seriousness of the dating relationship was tested additionally after selecting students who ha d a current dating partner. A total of 686 students had a current dating partner in the Korean subsample of psychologi cal abuse v ictims E ither having a steady or exclusive dating partner or hav ing an other format such as engaged, 12 T he result of logistic regression is th at being female increased the odds by 58%. A lso, the OLS regression failed to show the race difference, logistic regression finds that Hispanic students are less likely to be a victim of psychological abuse by their dating partner.

PAGE 82

82 married, or divorced failed to show a significant effect on victimization by psychological abuse compared to students having an occasional dati ng partner13. In sum, victimization models for the entire sample showed very significant results. Direct experience of child abuse in social learning theory and low self -control in self -control theory showed the very significant relationship to victimizatio n in physical and psychological abuse. Gender effect was additionally significant in victimization by dating violence Females were more likely to be victimized by their partner T he result s of the subsample for student s who ha d a dating partner indicat ed that ser i ous dating relationships affect ed victimization. O verall R2 has not changed a lot for the additional regression, however, the standardized coefficients (beta) decreased in the subsample model compared to the entire sample. T he result s of perpetration and victimization by dating violence in the U.S. were not investigated separately for non -perpetrators and perpetrators, or for non -victims and victims T he next section will investigate the models of dating violence for perpetrators and victims The following section will report the results of experience -specific models. F or example, perpetration of physical violence was tested for perpetrator s of physical violence, and victimization by physical violence was tested for victims of physica l abuse. P erpetration of physical and psychological violence: the subgroup of perpetrators only To determine significant factors among perpetrator groups, students who had ever perpetrated abuse against their partners were selected. Table 4 9 shows the re sults of the subsample of perpetrators of physical violence in the U.S. Among 228 students who had ever perpetrated physical abuse against their partner, only low self control was seen to affect 13 Additional result of O LS regression for subsample showed the influence of family structure. Being raised by adoptive parents or relatives increased the odds of being a victim by psychological abuse compared to being raised by two biological parents among students how have a dat ing partner.

PAGE 83

83 perpetration of physical abuse as a theoretical variable ( b= .174, b=.000) All three variables of social learning failed to predict the perpetration of physical violence statistically. H owever, direct experience of child abuse and attachment to father showed moderate predictable power under .10 criteria of pvalue. D irect experience of child abuse showed a positive relation to perpetrator s of physical violence (b=.360, p=.062). Attachment to father showed a negative relation to perpetrator of physical violence ( b= -.147, p=.074). Unsecure attachment to father incr eased the likelihood of perpetration of physical violence among physical perpetrators. W e rather found that gender (b=2.940, p=.001) race (b=2.160, p=.039) and risky sexual behavior (b=.223, p=.015) were more significant factors among perpetrator groups. That is, students having higher scores on risky sexual behavior and older students were more likely to engage in perpetration of physical violence. Females were significantly more likely to engage in perpetration of physical violence compared to male stud ents, and black students were more likely to be perpetrators compared to white students. Although previous OLS regression for the entire sample showed that having a serious dating partner was a significant factor, the variable was not significant among physical perpetrators. The most significant factor among those variables was low self -control, and the Beta value is .276. None of the other factors including witness of inter parental violence, age, family structure or living on or off campus affected perpetration of physical violence for physical perpetrators. Overall model has good fit (F=3.791, p=.000) and R2 is relatively high ( R2=.221). The perpetrator of psychological abuse also shows similar patterns to the perpetrator of physical violence. Tabl e 4 9 also shows the results of perpetration of psychological abuse among psychological perpetrators There was a relatively high number (N=591) of perpetration of psychological abuse than physical perpetrators Direct experience of child abuse, low self -

PAGE 84

84 c ontrol, gender, race, and risky sexual behavior influenced the perpetration of psychological abuse for perpetrators of psychological dating violence. Among students who had ever perpetrated psychological abuse against their partners, those students who had experienced direct violence by parents in childhood (b=.446, p=.000) and had low self -control ( b= .107, p=.000) increased the likelihood of perpetration. Females (b=1.393, p =.003) were more frequently perpetrators of psychological abuse against their pa rtners, and risky sexual behavior ( b=.301, p=.000) is still highly correlated to perpetration of psychological abuse. The interesting different factor is the effect of the other racial variable. Although black students showed stronger net effect on physi cal assault results, the other race category (including Pacific Islander, American Indian or Alaskan Native and other) (b=.096, p=.017) was more involved in perpetration of psychological abuse than were white students. None of the other variables, inclu ding attachment to parents witnessing inter -parental violence, age, family structure, dating involvement or living on or off -campus affected perpetration of physical violence among physical perpetrators. Overall model has good fit (F=.5.593, p=.000) and R2 is 164. This section showed the results of perpetration of dating violence for perpetrators. Theoretical factors such as direct experience of child abuse and low self -control showed a commonly significant effect on perpetration of dating violence in bo th physical violence and psychological abuse. F emale students were more likely to perpetrate physically and psychologically than male students Risky sexual behavior wa s also a significant factor for both forms of perpetration. Unsecure attachment to father was the significant effect for physical perpetrators but not for psychological abusers. A lthough race difference was varie d by the form of perpetration white students we re less likely to be a perpetrator than Hispanic (physical viol ence) and the other racial group ( psychological abuse). Interestingly, dating involvement

PAGE 85

85 was not significant for both models. In addition, although additional OLS regression was tested for student s who did have a current dating partner due to the tauto logical issue, the result failed to report a significant effect of the seriousness of the dating relationship. T he following section will cover the result s conducted on the subsample which included only students who ha d ever experienced victimization by dating violence. Victimization by physical and psychological violence: the subgroup of victims only Table 4 10 shows the OLS regression result selected only for victims of physical violence. Within the U.S. study group, 218 students reported suffering physical violence by their dating partner. Low s elf -control, attachment to father, family structure and risky sexual behavior showed significant predictive power for victimization by physical vio lence among victims of physical violence. Among them, risky sexual behavior ( b=.263, p=. 000) most increased the likelihood of victimization by physical violence. Interestingly, the social learning variable lost the power to predict S elf -control and some part of attachment theory showed a significant relationship with victimization by physical violence. Among 21 8 students, lower self -control (b= .107, p=.042) and insecure attachment to the father (b= -.190, p=.042) affected victimization by physical violenc e. However, attachment to mother, gender, age, race, family structure, and living on -campus did not show significant relationships with victimization by physical violence. Although p-value is slightly high (p=.058), students who were raised by single parents were more victimized than those raised by two biological parents. Overall model has good fit (F=2.397, p=.01) and R2 is relatively high ( R2=.196). Table 4 1 0 also shows the result of victimization by psychological abuse among victims of psychological abuse. Of the U.S. study group, 511 students reported that they were victims of psychological abuse. L ow self -control (b= .056, p=.014) and risky sexual behavior (b=.330, p=.000) were significantly predictive for victimization by psychological abuse. Although p -value

PAGE 86

86 of attachment to father is slightly higher than 05, it showed negative relation to psychological victimization for victims of psychological abuse ( b= -.106, p=.057). Other variables such as direct experience of child abuse, witness of inter -par ental violence, attachment to mother, gender, age, race, family structure, living on or offcampus and dating involvement did not play a significant role in the model of victimization by psychological abuse among the subsample of psychological abuse. O ver all model has good fit (F=3.683 p=.000) and R2 is moderate (R2=.129). In addition, although additional OLS regression was tested for student s who did have a current dating partner due to the tautological issue, the result s failed to report a significant effect of the seriousness of a dating relationship. In sum, victimization by the subsample who ha d experienced dating violence showed different result s compared to the entire student group. Low self -control, unsecure attachment to father, and high ly ris ky sexual behavior all increased the likelihood to be a victim by psychological abuse among students who were ever victimized by psychological abuse by their dating partner. D irect experience of child abuse failed show a significant effect in both victimiz ation model s n or was g ender effect found among victims. Moreover, unlike previous the victimization model for the entire sample, subsample result s reported that dating involvement did not affect physical or psychological victimization even among the additional group in a current dating relationship. Although the net effect of gender was not found very frequently, five models of American sample indicated the gender effect on dating violence. The next section will investigate how differently or similarly da ting violence is explained by gender in American sample based on the previous model that indicated gender effects.

PAGE 87

87 Dating Violence by Gender in the U.S. Gender effect was found in the perpetration of physical violence, perpetration of psychological abuse and victimization by psychological abuse for the entire model. Also, gender effect was found for the students group for at least 1 point on dating violence perpetratio n both in physical violence and psychological abuse. Table 4 1 9 shows the result of for perpetration of physical violence by gender group in the U.S. According to the male group for the entire sample, the fathers violence against the mother was a signific ant factor in explaining the perpetration of physical violence. Self -control was also very significant, hence, those who had lower self -control were more likely to be a perpetrator of physical violence. Those students raised by adoptive parents or relative s were more likely to be involved in perpetration of physical violence than those raised by two biological parents, and being engaged, married or divorced increased the likelihood of perpetration of physical violence. Also, members of the other racial gr oup (including Pacific Islander, Native Indian and Alaska Natives) were more likely to become involved in physical violence. The female group showed slightly different patterns compared to the male group for perpetration of physical violence. Instead of th e indirect violence experience of parents, direct experience of child abuse was a significant factor as was low self -control. Female students having a steady dating partner, living on -campus, or having high scores on risky sexual behavior were all more lik ely to be a perpetrator of physical violence. Black female students were more likely to become involved in perpetration of physical violence compared to white female students. Table 4 2 0 shows the result of perpetration of psychological abuse by gender group. In the case of the male perpetration model of psychological abuse, instances of direct child abuse, low self -control, risky sexual behavior, or being black as a rule were factors more likely to predict psychological abuse against their dating partner The female perpetration model of psychological

PAGE 88

88 abuse showed similar predictable variables. Similar to the male model, direct experience of child abuse, low self -control, and risky sexual behavior played a significant role to predict perpetration of psychological abuse. In addition, unsecure attachment to the mother as well as having a steady dating partner, family structure and living on -campus were new predictable factors for the female perpetration model of psychological abuse. That is, those female stu dents who formed unsecure attachments to the mother, had a steady or exclusive dating partner, or lived on -campus were more likely to be perpetrators of psychological abuse. Also, female students raised by stepparents were more likely to become perpetrator s of psychological abuse compared to those raised by two biological parents. There was no racial difference. The victimization model also showed gender differences ( Table 4 21). In the case of victimization by psychological abuse, male groups were affected by direct experience of child abuse, low self -control, and risky sexual behavior. Also, black males were more likely to be victimized psychologically by their dating partner compared to white male students. In the female victimization model of psychologi cal abuse, variables related to social learning could not explain the victimization. Rather, low self -control, having a steady or exclusive dating partner and engaging in risky sexual behavior were significantly related to female victimization by psycholog ical abuse. The gender difference was also found in student groups of perpetrators and victims. Table 4 22 shows the result of perpetration of physical violence among perpetrator group by gender. With regard to the male group of physical violence perpetra tors, neither social learning variables nor self -control explained the relationship with perpetration of physical violence. Interestingly, only family structure and race could predict the perpetration of physical violence. Those students who were raised by adoptive parents or relatives were more often involved in perpetration of

PAGE 89

89 physical violence. Asian male students were more likely to become involved in perpetration of physical violence compared to white male students. However, the female perpetrator grou p showed that direct experience of child abuse, low self -control, age, family structure, living on campus, risky sexual behavior and race were significant factors to predict perpetration of physical abuse. Those who had direct experience of child abuse, lower scores on self -control, as well as older female students, black female students, female students who were raised by adoptive or relatives, lived on-campus, or had risky sexual behavior were all more likely to become involved in perpetration of physical violence. Although the previous regression model for perpetrators and victims of psychological abuse shows the gender difference, the significant factors of perpetration of psychological abuse were direct experience of child abuse, low self -control, and r isky sexual behavior in both females and males. I have investigated the predictive factors of perpetration and victimization of dating violence in American society. In brief, unlike Korean model that indicated the age and living on campus effect on dating violence, age and campus effect were not found in American samples. Social learning variables such as direct experience of child abuse and self -control are consistently significant factors for dating violence in the U.S. In addition, the effect of unsecure attachment to father was found among students who have ever experienced dating violence as a perpetrator or a victim although it was not found among entire American students Gender effect among American students was clearer than Korea students Female s tudents are more likely to perpetrate and to be victimized than male students in the U.S. There were partial race effects on dating violence in the U.S. Being a black student increased the likelihood of perpetration of physical violence compared to white s tudent and being a racial minority such as Pacific Islander

PAGE 90

90 or native Indian increased the likelihood of victimization by physical violence and perpetration of psychological abuse among perpetrators. H owever the race effect is open to the question due to small numbers. I have investigated the predictive factors of perpetration and victimization of dating violence in both societies separately so far. Now I will investigate net effect of independent vari ables in the full model using combined data. T he analytical method s are identical to the previous separate d country models i.e., OLS regression for the entire group was run first and OLS regression for the subsample who had experienced dating violence was run later. F ull M odel Results (including U.S. and Korea) T he following section reports the results of the full model. Korean and U.S. data were combined. T he analytical method s are identical to the previous separate d country models i.e., OLS regression for the entire group was run first and OLS regression for the subsample who had experienced dating violence was run later In addition, logistic regression for all students and OLS regression for students in a dating relationship were tested; however, OLS regression will be mainly reported and only significant differences of logistic regression will be reported. Scales of each of four dependent variable s, child abuse, were re -created by factor analysis and other variables remained the same to previous OLS regression models. Instead of excluding the race variable, the country variable was added for full models. Descriptive Statistics for the Full Sample Table 4 1 1 presents the descriptive statistics for the combined full sample The sample size of collected data is 2,695 students from Korea and the U.S : 1 399 Korean (51.1%) and 1,296 American (48.9%) students participated in this survey. Among them, 1,078 students were male (40%), and 1,617 students (60%) were female. Among the partic ipants, 656 students ( 24.4%) were 18 years of age, 653 students (2 4.2 %) were 19, 489 students ( 18%) were 20, and 908

PAGE 91

91 students ( 33.7%) were 21 years of age or older. 554 students ( 20.6 %) in the full model answered they lived in on -campus residence halls, al most 2 135 (79.2%) students lived off -campus. Nearly all students in the sample ( 2 283, or 84.7%) answered that they were raised by two biological parents, 248 students ( 9.2 %) stated that they were raised by single parents, 117 students ( 4.3 %) were raised b y stepparents and 47 students (1. 7 %) were raised by adoptive parents, other relatives and others. A total of 2 695 students responded to the survey question regarding current dating relationship status. N ear half ( 45.5%) did not have a current dating partner at the time. The next largest group, 992 students (3 4.2 %) answered they had a steady or exclusive dating partner. 211 students (16.6 %) reported they were dating a partner occasionally. The balance (3. 9 %) answered t hat they were engaged, married or divorced, or had other forms of dating. These 105 students were merged into other in the statistical analysis. Three questions addressed risky sexual behavior such as age upon first sexual intercourse, a total number of sexual partners and a total number of sexual partners during the last three months. T he risky sexual behavior ranged from 0 to 18. H igher score mean t more risky sexual behavior Child abuse was measured using 8 items, the scale of child abuse ranged from 0 to 8 Less than half of the survey group, or 1 166 students, had a 0 score, meaning they had never experienced any abusive behavior by their parents in childhood, and the mean score was 1. 39. In questions regarding parents violence observed during child hood, 400 students ( 14.95%) answered they had witnessed their fathers violence against their mother, and 216 students (8%) answered they had witnessed their mothers violence against the father. Self control was measured by a total of 23 items used by G rasmick et al. (1993). Students responded to a four answer scale (1=strongly agree; 4=strongly disagree) and a low score indicated low self -control. Self -control scores ranged from 23 to 115, and the mean score of self -

PAGE 92

92 control is 6 4.92. Attachment to pare nts was divided into two measures: attachment to mother and attachment to father. The participants responded to four questions on a 5 -point scale: (1) how often do you trust your mother/female guardian? (1=never; 5= always ). Scores of attachment to mother ranged from 4 to 24, and the mean score of attachment to mother is 17.19. Scores of attachment to father ranged from 4 to 24, and the mean score of attachment to father is 15. 70. The survey addressed four major forms of dating violence: perpetration of p hysical violence, perpetration of psychological abuse, victimization by physical violence and victimization by psychological abuse. Perpetration of dating violence was measured using two forms. Perpetration of physical violence measures ranged from 0 to 36 The mean score of perpetration of physical violence is 92. The large majority, 2,048 students (80.6 %) had not perpetrated physical violence against their partner in the past year, yet 494 students (1 9.4 %) did perpetrate physical violence against their d ating partner in the past year. Perpetration of psychological abuse measures ranged from 0 to 28 and the mean score of psychological abuse 2.60. Roughly more than half (56.6 %) of student s had never perpetrated psychological abuse against their partner i n the past year, and 1 066 students ( 43.4%) did perpetrated psychological abuse against their partner in the past year. Victimization by dating violence was also measured using two forms. Victimization by physical violence measures ranged from 0 to 36, and the mean score is 73. A not insignificant number of students ( 15.3%) reported experiencing physical violence by their partner in the past year. Victimization by psychological abuse measures ranged from 0 to 28 and the mean score of victimization by psy chological abuse is 2.49. Less than half (40.8 %) of students were victimized by their dating partner in the past year. Consistently, scores

PAGE 93

93 of perpetration of and victimization by psychological abuse are higher than perpetration of and victimization by physical violence. Results of Multivariate Analyses for Full Model P hysical and psychological perpetration of dating violence in the full model OLS regression analysis w as run to find the net effects of variables among the entire student population from bo th groups regardless of the experience of dating violence. Among all students, regardless of their experience of perpetration or victimization by dating violence, this OLS regression will test significant factor s that increase the likelihood of perpetratio n of or victimization by dating violence. Although logistic regression was run for full model, only significant difference will be reported. T able 4 1 2 shows the result of perpetration of physical violence for combined full model. A total of 2,456 samples were used to test the perpetration of psychological abuse in the full model. Overall model fit is very significant (F=16.886) and R2=.105. OLS results predicting the perpetration of physical violence show that direct experience, lower self -control, attach ment to father, gender, a serious dating partner, and risky sexual behavior are commonly significant factors to predict perpetration of physical violence14. T hat is, more direct experience of child abuse increased the likelihood of being a physical perpetrator ( b=.192, p=. 000). Having more risky sexual behavior ( b=.100, p=.000) and lower score of self -control (b= -.052, p=.000) are significantly predictive of a higher likelihood of perpetration of physical violence for both Korean and U.S. students Unsecure attachment to father also affect perpetration of physical violence. S tudents who had lower score on attachment to father were more likely to perpetrate 14 A ccording to our result s of logistic regression of perpetration of dating violence, direct experience of child abuse, low self control, gender, living in Korea versus living in the U.S., having a serious dating partner and risky sexual behavior are related to perpetration of dating violence for college students

PAGE 94

94 their dating partner physically (b= .030, p=.048). Females w ere more likely to be perpetrator s against their dating partner of psychological abuse than are male students ( b=.704, p=.000)15. U.S. students w ere more likely to be a perpetrator of physical violence than Korean students (b=.459, p=.004).16 In addition, interestingly, the full model combini ng both the Korean and U.S. group shows that the students who were raised by stepparents w ere more likely become involved in perpetration of physical violence compared to students who were raised by two biological parents (b=. 652, p=.037). Although having a dating partner showed the significant effect on perpetration of physical violence compared to not having a dating partner, due to tautological issues, it required additional analysis. 1 332 students who had a dating partner were selected and having an occasional dating partner is a reference group. Having other format of dating increased the likelihood of being a perpetrator of physical violence compared to having an occasional dating partner. O ther variables such as witness of inter parental violence, attachment to mother, age, and living on or off-campus failed to show the significant effect on perpetration of physical violence. P erpetration of psychological abuse was tested similarly ( Table 4 1 2). A total of 2,373 samples were used to test the perpetration of psychological abuse in the full model. T able 4 1 2 also shows the result of perpetration of psychological abuse in the full model. T he variance explained by this model is 25.3% and model fits well (F=46.95). OLS result s predicting the perpetration by psychological abuse show more direct experience of child abuse, lower self -control, being female, being U.S. students having a serious dating partner, and engaging in risky sexual behavior are related to perp etration of psychological 15 Logistic regression reported that being female increased the odds of perpetration of physical violence by 88.5%. 16 L ogistic regression of physical perpetration reported that being U.S. student increased the odds of being a perpetrator of physical violence by 58.4%.

PAGE 95

95 abuse. Students who experienced child abuse by parents w ere more likely to be a perpetrator of psychological dating violence (b=.288, p=.000). Lower self -control consistently show ed significant power to explain the perpetration of psychological abuse (b= .068, p=.000). Female students w ere more likely to perpetrate psychological abuse than male students ( b=1.148, p=.000)17. Also, U.S. students w ere more likely to perpetrate psychological dating violence than Korean students ( b=2.144, p=.000)18. In addition being raised by stepparents increased the likelihood of perpetration of psychological abuse compared to students raised by two biological parents. W itness of inter -parental violence, attachment to parents, age, and living on or offcampus was not significant factors Although having a dating partner showed the significant effect on perpetration of physical violence compared to not having a dating partner, due to tautological issues, it required additional analysis. 1,301 stude nts who had a dating partner were selected and having an occasional dating partner is a reference group. Having other format of dating or steady/ exclusive dating partner increased the likelihood of being a perpetrator of psychological abuse compared to h aving an occasional dating partner in the subsample model. Physical and psychological victimization by dating partner in the full model Victimization by physical violence was tested similarly. A total of 2,463 samples were used to test the perpetration of physical abuse in the full model. T able 4 13 shows the result of victimization by physical violence in the full model. T he variance explained by this model is 7.8% and model fits well (F=12.123). OLS results predicting victimization by p hysical violence s howed the direct experience of child abuse, lower self -control, having a serious dating partner, engag ing in risky sexual 17 T he factor of gender, especially among female students increased the odds of being a perpetrator of psychological abuse by 76.5%. 18 T he factor of country, especially among U.S. students increased the odds of being a perpetrator of psychological abuse by 140%.

PAGE 96

96 behavior and country of residence we re higher risk factors for being a victim. Students who ha d experienced abuse in childhood w ere more likely to be a victim by physical violence by their dating partner (b=.164, p=.000). Lower self -control influenced the likelihood of victimization by physical violence ( b= -.037, b=.039). C ountry of residence also showed a significant effect on victimi zation by physical violence. St udents in the U.S. w ere more likely to be victimized physically by their dating partner ( b=.377, p=.008)19. Gender difference was not found in results on victimization by physical violence. H owever, an attachment effect was re vealed. A ttachment to the mother showed a negative relationship with victimization by physical violence ( b= .037, p=.000). That is, unsecured attachment to the mother affect ed victimization by physical violence. Although having a dating partner showed a si gnificant effect on perpetration of physical violence compared to not having a dating partner, due to tautological issues, it required additional analysis. 1 337 students who had a dating partner were selected and having an occasional dating partner wa s a reference group. Having other format of dating, such as engaged, married, or divorced, increased the likelihood of be ing a victim by physical violence compared to having an occasional dating partner. O ther variables such as family structure age, wit ness of inter -parental violence, or living on or off campus failed to show a significant effect on victimization by physical violence. Victimization by psychological abuse was tested similarly. A total of 2 3 84 samples were used to test the victimization by psychological abuse in the full model. T able 4 13 shows the result s of victimization by psychological abuse in the full model. Overall model fits well (F=44.288) and used variables predicted 24.1% variance of victimization by psychological abuse. 19 C ountry effect increased the odds of victimization by physical violence by 70%.

PAGE 97

97 Victimization by psychological abuse showed results similar to the previous three variables of dating violence. Victimization by psychological abuse is predicted by child abuse, low self control, gender, having a serious dating partner, being among the U.S. stu dents group, and risky sexual behavior D irect experience of child abuse showed a positive relation to victimization by psychological abuse ( b=.259, p=.000), low self -control showed a negative relation to victimization by psychological abuse (b= .063, p=.00). Country of residence wa s a very significant factor. Being among the U.S. students group increased the likelihood of victimization by psychological abuse over Korean students (b=1.887, p=.000)20. F emale students w ere more likely to b e a victim by psychological abuse by their dating partner than were male students (b=.865, p=.000)21 In addition, family structure showed a significant relationship with victimization by psychological abuse. That is, students who were raised by stepparents were more likely to be a victim by psychological abuse than students who were raised by two biological parents ( b=1.024, p=.013). Although having a dating partner showed a significant effect on victimization by psychological abuse compared to not having a dating partner, due to tautological issues, it required additional analysis. A total of 1 301 students who had a dating partner were selected and having an occasional dating partner wa s used as a reference group. Reporting another form of dating (such as engaged, married, or divorced) increased the likelihood of be ing a victim by psychological abuse compared to having an occasional dating partner. In sum, victimization models for the full model sample indicated tha t direct experience of child abuse, low self -control country of residence, and risky sexual behavior were commonly 20 C ountry effect, U.S. students incr eased the odd s of victimization by psychological abuse by 48%. 21 Gender effect, being female, increased the odd s of victimization by psychological abuse by 63%.

PAGE 98

98 significant in both forms of dating violence. Child abuse showed a positive relationship and low self -control showed a negative relationship with physical and psychological victimization. U nsecure attachment to mother was significant factor for victimization by physical violence, and gender effect found at only psychological victimization F emales w ere more likely to be victimized psychologica lly. I n addition, being raised by stepparents also increased the likelihood of being a victim by psychological abuse compared to being raised by two biological parent s T he result of perpetration and victimization by dating violence usin g combined data ha s not been investigated separately for non -perpetrators and perpetrators, and non-victims and victims N ow the next section will investigate the models of dating violence for perpetrators and victims The following section will report the result of experie nce -specific models. F or example, perpetration of physical violence was tested for perpetrator s of physical violence, and victimization by physical violence was tested for victims of physical abuse. P erpetration of physical and psychological violence amon g perpetrators To determine significant factors among perpetrator groups, the students who had ever perpetrated abuse against their partners were selected. Table 4 14 shows the results of the subsample of perpetrators of physical violence in the full model. Among 480 students who had ever perpetrated physical violence against their partner, t he result shows that direct experience of child abuse, self control and attachment to father were seen to affect perpetration of physical violence as a theoretical variable T he variance explained by this model is 14.1% and model fits well (F=4.475). Students who had experienced child abuse in childhood w ere more likely to be a perpetrator of physical violence (b=.446, p=.000). Lower self -control influenced on per petration of physical violence among physical perpetrators (b= -.144, b=.000). Country of residence failed to show a significant effect on perpetration of physical violence among perpetrators. Gender

PAGE 99

99 difference was found in results on perpetration of physic al violence. Among perpetrators, females w ere more likely to perpetrate physical violence against their partners than were male students (b=1.363, p=.019). In addition, an attachment effect was revealed. A ttachment to the father showed a negative relations hip with victimization by physical violence ( b= .116, p=.057). That is, unsecured attachment to the father increased the likelihood of perpetration of physical violence. Although having a dating partner showed a significant effect on perpetration of physic al violence compared to not having a dating partner, due to tautological issues, it required additional analysis. O nly 367 students who had a dating partner were selected and having an occasional dating partner wa s a reference group. Having other format of dating or steady or exclusive dating partner increased the likelihood of being a perpetrator of physical violence compared to having occasional dating partner. Other variables such as family structure age, witness of inter -parental violenc e, living on or off campus and risky sexual behavior failed to show a significant effect on perpetration of physical violence among physical perpetrators. OLS Regression analysis for the group of perpetrator s of physical violence was run to find the signi ficant factor s for perpetrator s of physical violence. Among the perpetrator s of physical violence, the risk factor s increasing the likelihood of perpetration of physical violence were child abuse experience, low self -control, gender, and having a serious d ating partner. T hat is, among students who ha d ever perpetrated physical violence against their partner, more experience of direct child abuse increased the likelihood of being a perpetrator of physical violence. If students ha d lower self -control, they w e re more likely to be a perpetrator compared to students having higher self -control. H aving a serious dating relationship also increased the likelihood compared to students having no dating partner. A lso, females w ere more likely to be perpetrator s than mal e students even among the perpetrator group. N ational factor s failed to show the significance for

PAGE 100

100 perpetration of dating violence for the group of perpetrator s A lthough it is not statistically strongly significant under .05, attachment to the father showed a negative relationship with perpetration of physical violence among the perpetrator group. T hose students with unsecure attachment to the father were more likely to perpetrate physical violence against their dating partner (p=.057). Table 4 14 sho ws the results for the subsample of perpetrators of psychological abuse in the full model. Among 1 030 students who had ever perpetrated psychological abuse against their partner, t he result s indicated that direct experience of child abuse, self control and country, gender, and risky sexual behavior were seen to affect perpetration of psychological abuse Students who had experienced child abuse in childhood w ere more likely to be a perpetrator of psychological abuse (b=.364, p=.000). Lower self -control influenced on perpetration of physical violence among physical perpetrators ( b= .103, b=.000). Country of residence show ed a significant effect on perpetration of psychological abuse among perpetrators. U.S. students w ere more likely to perpetrate physical violence against their dating partner than were Korean students (b=3.209, p=.000). Gender difference was found in results on perpetration of psychological abuse. Among perpetrators, females w ere more likely to perpetrate their dating partner psychologically than male students (b=.904, p=.002). Higher risky sexual behavior wa s related to a higher likelihood of perpetration of psychological abuse among psychological abusers ( b=8.24, p=.000). Although having a dating partner showed a significant effect on perpetration of psychological abuse compared to not having a dating partner, due to tautological issues, it required additional analysis. O nly 762 students who had a dating partner were selected and having an occasional dating partner wa s a ref erence group. Having another format of dating

PAGE 101

101 or a steady or exclusive dating partner increased the likelihood of being a perpetrator of psychological abuse compared to having an occasional dating partner. Other variables such as family structure age, wit ness of inter -parental violence, living on or off campus and dating involvement failed to show t significant effect on perpetration of psychological abuse among psychological perpetrators. A lthough it showed similar result s to the previous model of perpe tration of physical violence, the national factor and risky sexual behavior showed a more significant relationship with perpetration of psychological abuse compared to that of having a serious dating partner M ore direct experience of child abuse, lower se lf -control, being female, U.S. students, and engag ing in risky sexual behavior we re significantly related to perpetration of psychological abuse. Victimization by physical and psychological dating violence among victims Table 4 15 shows the results of the subsample of victims of physical violence in the full model. OLS regression for victimization by physical violence has 376 samples. T he variance explained by this model is 11.2% and model fits well (F=2.648). A lthough it showed similar result s to the pr evious model of perpetration of physical violence, OLS regression for victims of physical violence indicated that only child abuse, low self -control and having a serious dating partner affect victimization by physical violence. Among students who ha d ever experienced victimization by physical violence by their dating partner, students who ha d more experience of child abuse (b=.410, p=.012) w ere more likely to be victims than students with less experience of child abuse. Students with lower self -control w ere more likely to be victimized by physical violence by their partner H aving a more serious dating partner also increased the risk of victimization by physical violence. Like the previous model for the perpetrator of physical violence the national factor failed to show significance. A lthough it is not statistically strongly

PAGE 102

102 significant under .05, attachment to the father showed a negative relationship with perpetration of physical violence among victims of physical violence ( b= .136, p=.061). T hose student s with unsecure attachment to the father w ere more likely to be victim by their dating partners physically (p=.061). However, risky sexual behavior failed to show a significant effect on this model for victims. In addition none of the other variables, inc luding witness of inter parental violence, attachment to mother, country, gender, age, family structure or living on or off-campus were related to victimization by physical violence for victims of physical violence in the full model. Table 4 15 also shows the result of subsample of victims of psychological abuse in the full model. OLS regression for victimization by psychological violence has 974 samples. T he variance explained by this model is 20.7% and model fits well (F=14.660). OLS regression for victi ms by psychological abuse indicated child abuse, low self -control, U.S. students, engag ing in risky sexual behavior and having a serious dating partner affect victimization by psychological abuse. A mong students who ha d ever been victimized by psychologica l abuse by their dating partner, direct experience of child abuse increased the likelihood compared to students with less experience of child abuse ( b=.340, p=.000). Those students having lower self -control significantly increased the likelihood of victimization by psychological abuse than students with higher self -control ( b= .086, p=000). A lthough it is not statistically strongly significant under .05, attachment to the father showed the negative relationship with victimization b y psychological abuse among victims by psychological abuse. T hose students with unsecure attachment to the father w ere more likely to be psychologically victimized by their dating partners (b= -.066, p=.092). Serious dating relationship s increased the likel ihood of victimization by psychological abuse over other dating format s S tudents having higher score s on risky sexual

PAGE 103

103 behavior we re more likely to be a victim than other students with lower score on risky sexual behavior ( b=.285, p=.000). A lthough gender difference was not found, national difference was found. U.S. students w ere more likely to be victims than Korean students (b=.285, p=.000). None of the other factors including witness of inter parental violence, attachment to mother, gender, age, family s tructure, and living on or off -campus was significant to predict the victimization by psychological abuse for psychological dating violence victims in the full model. I have investigated significant factors for full model so far. I n short, the significance of direct experience of child abuse, self -control, and risky sexual behavior are consistent to separate U.S. and Korean models. Consistently, other variables of social learning such as witnessing of father s violence against mothe r, and witnessing of mother s violence against father failed the significance in the full models. F emale students are more likely to get involved in dating violence than male students in the full mode. U nsecure attachment to father affect the perpetration of physical violence, perpetration of physical violence among perpetrators, and victimization by psychological abuse among victims by psychological abuse. Unsecure attachment mother influenced victimization by physical violence in the full model. F amily structure also showed the significance in the psychological aspect of dating violence. B eing raised by step parents increased the likelihood of perpetration and victimization by psychological abuse. A s m entioned earlier, country of residence was a significant factor to predict dating violence in the full model. I f it is true that the country is a significant factor to explain dating violence in the full model, then the next question will be how differen tly the factors explain the dating violence. Unfortunately, although current OLS regression could explain the net effect of country, it is not allowed to compare the power of significant independent variables in both

PAGE 104

104 countries. The n ext section will inves tigate how differently or similarly each significant factor explains dating violence in both countries. In addition, gender difference will be investigated as far as the net effect of gender has been found on the previous models in each country. Coefficie nt Comparison Paternoster et al. ( 1998) suggested a formula (see Equation 3 1 ) for two -sample comparison. I t allows comparing whether independently obtained coefficients are significantly different U sing this formula, I will compare two countries to test how significant factors explain different ly in each country. I n addition, the recommended formula by P aternoster et al. (1998) will be used to compare gender difference. For comparison by country, regression coefficients obtained in the separate country model s has been compared to find differences in the predict ive power in each society. A fter running OLS regression in each country model for the entire student sample as well as the perpetrator and victims group, the commonly significant factor in each model were selected. Comparison by Country P erpetration of physical violence for the entire student model regardless of experience of dating violence ha d five common significant factor s: direct experience of child abuse, low self control, having a s teady dating partner, having an other dating partner format, and risky sexual behavior. A mong them, only low self -control showed different predictive power s in each countr y T hat is, low self -control ha d more predictive power on the perpetration of physical violence in the U.S. compared to its effect in Korea. Although I conducted coefficient comparison tests for coefficients for U.S. scale s and Korean scale s using two sample t -test, it can be argued that it is a problem to use the formula for two scales that are not comprised of the same items. M ost of scales of Korean and American samples have the same items, but one independent variable ( child abuse) and one dependent

PAGE 105

105 variable ( victimization by psychological abuse) ha ve different number of items. A private communication from one of authors of the article (Paternoster et al., 1998), Robert Brame indicates that the coefficient comparison in the case of scales with different items should not be made Therefore, compariso n results that included child abuse and victimization of psychological abuse will not reported in this study. However, other variables with the same scale items will be discussed in this chapter. Table 4 26 shows the significance of difference in OLS regression coefficients for Korean and American models. P erpetration of psychological abuse ha d six common factors : direct experience of child abuse, low self -control, gender, having a steady dating partner, having a n other format of d ating, and risky sexual behavior. It is very interesting that all factors predicted differently in each society D irect experience of child abuse is more predictive of the perpetration of psychological abuse in Korean society, however, the comparison is no t valid due to different number of items in both countries L ower self -control is more predictive of the perpetration of psychological abuse in U.S. society. Gender effect on perpetration of psychological abuse is larger in Korean society. Also, having a s erious dating partner and engaging in risky sexual behavior proved a higher predictive power on the perpetration of psychological abuse in Korea than in the U.S. Victimization by physical violence had four common factors : direct experience of child abuse, low self -control, having a steady dating partner, and having another format of dating. H owever, t test result is not significant. This means all four factors explain the victimization by physical violence similarly in both societies Victimization by psychological abuse has common ly six significant factors in each country. M ost of the factors failed to show different effect s in each country but risky sexual

PAGE 106

106 behavior showed a significant difference in both countries. S tudents risky sexual behavior af fects victimization by psychological abuse among Korean students more so than among U.S. students However, I had different number of item to measure victimization of psychological abuse in both samples, so this comparison may not be useful. Perpetration of physical violence for the perpetrator group had two common factors Low self -control on perpetration of physical violence among perpetrators played a similar role in both countries. H owever, direct experience of child abuse showed a stronger effect on t he perpetration of physical violence among the Korean perpetrator group. Victimization by psychological abuse among victims ha d two common factors low self control and risky sexual behavior. These factors failed however, to show differences in results b etween the two countries T hroughout this country comparison, the sharp different power of explanation of significant factors in each country is remarkable for perpetration of psychological abuse. The power of self -control is stronger in the U.S. and the power of social learning is stronger in Korea. Comparison by Gender Gender d ifference in Korea Gender difference was found in the perpetration of psychological abuse and victimization by psychological abuse for the entire sample, and victimization by psychological abuse for the victim group. Table 4 24 shows the significance of difference in OLS regression coefficients for male and female models in Korea. Child abuse was commonly found in both gender groups to predict perpetration of psychological abu se in Korea. H owever, it is hard to say the power of prediction is different (z score=.0714). Victimization by psychological abuse has four common factors between male students and female students such as child abuse, self -control, having a steady dating p artner and

PAGE 107

107 risky sexual behavior. Interestingly child abuse, self -control and risky sexual behavior are showing the significance of difference to predict the victimization by psychological abuse in both gender groups O f course, these variables are very s ignificant factors in both groups, the power of explanation is different Whereas child abuse (z=1.765) and self -control(z=1.785) were more powerful factor for male students to predict victimization by psychological abuse compared to female students risky sexual behavior was more powerful predictors for female students to predict victimization by psychological abuse. The variable of risky sexual behavior was found as a common factor in the model of victimization by psychological abuse among victims the po wer of explanation is not different in both gender groups. Gender d ifference in the U.S. While Korean gender differences are found only in the model of psychological abuse the gender difference in the American samples are found in the both form of physi cal and psychological abuse. Five models showed the gender difference in previous OLS regression models; perpetration of physical violence, perpetration of psychological abuse, victimization by psychological abuse, perpetration of physical violence among perpetrators and perpetration of psychological abuse among perpetrators. T able 4 25 is showing the significance of difference in OLS regression coefficients for male and female models in the U.S. perpetration of physical violence has self -control commonly i n both gender groups. Because the z -score (1.35) is below than 1. 96, low self -control affect similarly perpetration of physical violence for male and female students While perpetration of psychological abuse and perpetration of psychological abuse among perpetrators have three variables as common factors equally (child abuse, self -control, and risky sexual behavior), different power of each variable was not found. T hat is, experience of child abuse in childhood, lower self -control, and higher risky sexual behavior similarly increased the likelihood of perpetration of psychological abuse for male and female students Moreover,

PAGE 108

108 victimization by psychological abuse also failed the different power of significant variable. S elf control was the common significant factor in male and female group, the t test result improved no difference of the power of explanation. Only the model of perpetration of physical violence among perpetrators showed the gender difference in the U.S. models. T he both gender groups have fa mily structure (other format) as common factor to predict perpetration of physical violence among perpetrators. The effect of being raised by adoptive parents or relatives is more powerful to predict perpetration of physical violence ( perpetrator group) am ong male students than female students However it needs to take the small number of students who were raised by adoptive parents and relatives into consideration. T hen, there might be no factors that are explaining dating violence differently in the two gender group in the U.S. Throughout this gender comparison, I found that Korean gender difference is remarkable at victimization by psychological abuse. To predict victimization by psychological abuse, the power of child abuse and self -control was notable for male students and the power of risky sexual behavior is more notable for female students Other models did not indicate the different power of significant factors by gender. H owever, it is very remarkable that there was no significant power differenc e by American gender groups. The theoretical factors that were applied to this dissertation did not showed the different power by gender group to explain dating violence. Although the reason has not discussed, the possible explanation of equality of explan ation power by gender among American students is that most of criminological theories have been developed in western society If the measure of criminological theories could reflect the cultural different for Eastern society, I expect that the result of Korean might not be different from American r esult

PAGE 109

109 Table 4 1 Descriptive statistics for the Korean sample Variables Mean Median Mode Sd Min Max Dependent Variables Perpetration of p hysical v iolence .63 .00 .00 2.49 .00 32.00 Perpetration of p sychological v iolence 1.08 .00 .00 2.36 .00 2 4 .00 Victimization by p hysical v iolence .57 .00 .00 2.50 .00 36.00 Victimization by p sychological v iolence 1.80 .00 .00 3.62 .00 20.00 Independent Variables Victim of c hild a buse 1.52 1.00 .00 2.00 .00 9.00 Witnessed fathers violence .23 .00 .00 .42 .00 1.00 Witnessed mothers violence .085 .00 .00 .28 .00 1.00 Self Control s cale 61.67 61.00 59.00 9.01 23.00 115.00 Attachment to mother 16.63 17.00 20.00 3.35 4.00 24.00 Attachment to father 15.19 16.00 20.00 4.21 4.00 24.00 Control Variables Sex .55 1.00 1.00 .50 .00 1.00 Age 2.93 3.00 4.00 1.12 1.00 4.00 Class attending 2.14 2.00 1.00 1.05 1.00 5.00 Family structure 1.10 1.00 1.00 .45 1.00 4.00 Dating involvement 1.87 1.00 1.00 .97 1.00 4.00 Campus .95 1.00 1.00 .22 .00 1.00 Risky s exual behavior 2.21 .00 .00 3.60 .00 18.00 N=1399

PAGE 110

110 Table 4 2 OLS regression models for perpetration of physical violence (all students) in Korea *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Model A(perpetration of physical violence) N=1304 Model B(perpetration of psychological abuse) N=1278 B P B P Independent Variables Victim of c hild a buse .135** .107 .000 .163** .139 .000 Witnessed fathers violence .082 .014 .648 .106 .019 .514 Witnessed mothers violence .131 .014 .622 .055 .007 .816 Self c ontrol scale .037** .117 .000 .041** .142 .000 Attachment to mother .021 .028 .398 .039 .056 .078 Attachment to father .008 .013 .699 .006 .010 .753 Control Variables Sex 3.146 .029 .318 .361** .077 .006 Age .133* .059 .038 .137* .065 .018 Family structure Single parents .028 .081 .936 .754 .064 .015 Step parents .706 1.106 .269 .349 .016 .541 Others 1.118 1.926 .054 .220 .011 .671 Dating involvement Occasionally .465** .066 .020 .811** .124 .000 Steady .524** .100 .001 .774** .153 .000 O thers 1.851** .125 .000 1.240** .092 .001 Campus .539 .046 .090 .492 .011 .092 Risky s exual behavior .056** .079 .008 .114* .171 .000 Constant 3.501 ** .000 3.093 ** .000 F = 6.826 F = 13.985 p = .000 p = .000 R 2 = .078 R 2 = .151

PAGE 111

111 Table 4 3 OLS regression model for victimization by psychological abuse (all students) in Korea *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Model A( victimization by physical violence) N=1308 Model B ( victimization by psychological abuse) N=1275 B P B P Independent Variables Victim of c hild a buse .104** .083 .005 .199** .110 .000 Witnessed fathers violence .074 .012 .681 .225 .026 .364 Witnessed mothers violence .188 .021 .481 .001 .000 .998 Self c ontrol scale .033** .105 .000 .056** .124 .000 Attachment to mother .037 .049 .137 .011 .010 .742 Attachment to father .000 .001 .987 .034 .040 .211 Control Variables Sex .004 .001 .981 .820** .113 .000 Age .049 .022 .444 .265** .082 .003 Family structure Single parents .057 .004 .870 .884 .049 .062 Step parents .957 .041 .870 1.290 .093 .138 Others .002 .000 .997 .143 .005 .856 Dating involvement Occasionally .277 .040 .165 1.175** .116 .000 Steady .455** .084 .004 1.478** .191 .000 O thers 1.556** .107 .000 .691 .033 .215 Campus .787* .068 .014 .373** .082 .003 Risky s exual behavior .040 .057 .062 .221** .217 .000 Constant 3.598 ** .000 3.588 ** .000 F = 4.802 F = 15.536 p = .000 p = .000 R 2 = .0 56 R 2 = .165

PAGE 112

112 Table 4 4 OLS regression model for perpetration of physical violence and psychological abuse (s ubgroup of p erpetrators o nly) in Korea *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Model A(perpetration of physical violence) N=193 Model B(perpetration of psychological abuse) N=400 B P B P Independent Variables Victim of c hild a buse .295 .129 .107 .199** .142 .006 Witnessed fathers violence .043 .004 .962 .291 .043 .425 Witnessed mothers violence .476 .031 .690 .565 .056 .267 Self c ontrol scale .108* .183 .015 .079** .217 .000 Attachment to mother .013 .009 .911 .000 .000 .998 Attachment to father .055 .047 .580 .018 .024 .669 Control Variables Sex .242 .023 .765 .321 .052 .307 Age .727* .161 .032 .110 .039 .432 Family structure Single parents .713 .029 .687 1.770** .126 .009 Step parents .498 .015 .833 1.042 .045 .357 Others 2.157 .060 .415 2.359 .077 .110 Dating involvement Occasionally .171 .015 .865 .916* .120 .027 Steady 1.050 .100 .246 .665 .108 .052 O thers 6.328** .244 .001 3.321** .186 .000 Campus 2.092 .094 .199 1.616* .110 .022 Risky s exual behavior .078 .057 .453 .061 .076 .153 Constant 14.93 ** .000 8.20 ** .000 F = 2 312 F = 4.874 p = .004 p = .000 R 2 = 174 R 2 = 169

PAGE 113

113 Table 4 5 OLS regression model for victimization by physical violence and psychological abuse (s ubgroup of victims o nly) in Korea Model A(victim of physical violence) N=158 Model B (victim of psychological abuse) N=419 B P B P Independent Variables Victim of c hild a buse .395 .162 .082 .208* .102 .045 Witnessed fathers violence 1.909 .156 .102 .693 .068 .191 Witnessed mothers violence .802 .049 .582 .257 .018 .728 Self c ontrol scale .089 .138 .090 .061* .113 .022 Attachment to mother .032 .022 .816 .016 .012 .820 Attachment to father .071 .055 .557 .134* .124 .024 Control Variables Sex 1.737 .150 .085 .992* .111 .030 Age .185 .037 .657 .317 .079 .106 Family structure Single parents 2.305 .078 .323 1.618 .076 .108 Step parents 1.260 .039 .631. 1.387 .046 .335 Others 2.694 .065 .443 .420 .010 .825 Dating involvement Occasionally .609 .045 .623 .538 .049 .364 Steady .738 .065 .479 .971* .110 .047 O thers 8.019** .271 .001 3.113* .097 .046 Campus 5.030* .206 .013 1.463 .067 .160 Risky s exual behavior .136 .092 .289 .250** .215 .000 Constant 17.498 .000 8.863** .000 F = 2.205 F = 4.342 p = .0 07 p = .000 R 2 = 200 R 2 = .147 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 114

114 Table 4 6 Descriptive statistics for the American sample Variables Mean Median Mode Sd Min Max Dependent Variable s Perpetration of p hysical v iolence 1 25 .00 .00 3.67 .00 3 6 .00 Perpetration of p sychological v iolence 3.83 1 .00 .00 5 22 .00 2 4 .00 Victimization by p hysical v iolence 89 .00 .00 3 03 .00 36.00 Victimization by p sychological v iolence 3 .0 4 .00 .00 4.73 .00 2 4 .00 Independent Variables Victim of c hild a buse 1.5 0 1.00 .00 1.82 .00 7 .00 Witnessed fathers violence 065 .00 .00 .2 4 .00 1.00 Witnessed mothers violence .0 75 .00 .00 .2 6 .00 1.00 Self c ontrol scale 6 8 43 6 8 .00 68 .00 10.77 23.00 115.00 Attachment to mother 1 7.79 1 9 .00 20.00 3. 23 4.00 24.00 Attachment to father 1 6.28 1 8 .00 20.00 4. 93 4.00 24 .00 Control Variables Sex 66 1.00 1.00 47 .00 1.00 Age 2. 28 2 .00 1 .00 1. 53 1.00 4.00 Class attending 2. 25 2.00 1.00 1. 11 1.00 5.00 Family structure 1. 36 1.00 1.00 71 1.00 4.00 Dating involvement 2 .07 2 .00 1.00 .97 1.00 4.00 Campus 62 1.00 1.00 48 .00 1.00 Risky s exual behavior 4.09 1 .00 1 .00 4.07 .00 18.00 N=1 296

PAGE 115

115 Table 4 7 OLS regression models for perpetration of physical violence (all students) in the U.S. *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating Model A(perpetration of physical violence) N=1158 Model B(perpetration of psychological abuse) N=1111 B P B P Independent Variables Victim of c hild a buse .208 ** .101 .001 .389 ** .132 .000 Witnessed fathers violence .001 .000 .998 .407 .019 .534 Witnessed mothers violence .404 .029 .367 .568 .029 .355 Self c ontrol scale .068 ** .172 .000 .086 ** .155 .000 Attachment to mother .016 .013 .655 .045 .027 .355 Attachment to father .036 .048 .125 .034 .030 .326 Control Variables Sex 1.25 ** .160 .000 1.636 ** .148 .000 Age .053 .016 .618 .083 .018 .563 Race Hispanic .067 .006 .827 .580 .039 .162 Black .742 .066 .029 .529 .033 .248 Asian .459 .027 .345 .992 .042 .127 Other .912 .040 .155 .765 .024 .375 Family structure Single parents .018 .002 .955 .577 .038 .191 Step parents .531 .038 .186 .951 .047 .084 Others .163 .006 .822 .392 .011 .690 Dating involvement Occasionally .254 .026 .406 .670 .048 .108 Steady .920 ** .121 .000 1.781 ** .167 .000 O thers 1.692 ** .089 .003 2.573 ** .099 .000 Campus .368 .048 .133 .256 .024 .441 Risky s exual behavior .124 ** .133 .000 .382 ** .294 .000 Constant 4.060 ** .000 5.146 ** .000 F = 8.716 F = 16.549 p = .000 p = .000 R 2 = 133 R 2 = 233

PAGE 116

116 Table 4 8 OLS regression models for victimization (all students) in the U.S. *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating Model A(victimization by physical violence) N=1161 Model B(victimization by psychological abuse) N=1118 B P B P Independent Variables Victim of c hild a buse .188 ** .109 .001 .296 ** .111 .000 Witnessed fathers violence .102 .008 .804 .219 .011 .723 Witnessed mothers violence .281 .024 .447 .250 .014 .654 Self c ontrol scale .041 ** .124 .000 .056 ** .113 .000 Attachment to mother .045 .046 .134 .015 .010 .736 Attachment to father .028 .045 .156 .022 .021 .491 Control Variables Sex .334 .052 .078 .720 .072 .011 Age .066 .025 .457 .050 .012 .712 Race Hispanic .117 .013 .651 .297 .022 .440 Black .163 .017 .564 .590 .041 .160 Asian .009 .001 .981 .341 .015 .585 Other 1.118 .058 .043 .814 .027 .331 Family structure Single parents .276 .031 .311 .370 .027 .363 Step parents .442 .038 .185 .460 .026 .361 Others .032 .001 .959 .700 .021 .453 Dating involvement Occasionally .185 .023 .468 .915 .073 .018 Steady .470 .075 .023 1.089 ** .112 .000 O thers 1.651 ** .102 .001 1.327 .056 .052 Campus .090 .014 .659 .017 .002 .955 Risky s exual behavior .120 ** .156 .000 .383 ** .326 .000 Constant 3.594 ** .000 3.871 ** .002 F = 6.828 F = 13.32 9 p = .000 p = .000 R 2 = 107 R 2 = 195

PAGE 117

117 Table 4 9 OLS regression model for perpetrators (subgroup of perpetrators only) in the U.S. *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating Model A(perpetrator of physical violence) N=228 Model B(perpetrator of psychological abuse) N=591 B P B P Independent Variables Victim of c hild a buse .360 .118 .062 .446 ** .157 .000 Witnessed fathers violence .532 .025 .394 .662 .034 .439 Witnessed mothers violence .843 .039 .542 .050 .003 .949 Self c ontrol scale .174 ** .276 .000 .107 ** .200 .000 Attachment to mother .041 .025 .683 .056 .036 .399 Attachment to father .147 .115 .074 .046 .040 .367 Control Variables Sex 2.940 ** .205 .001 1.393 ** .122 .003 Age .599 .117 .101 .144 .033 .485 Race Hispanic .759 .043 .457 .228 .015 .711 Black 2.160 .128 .039 .356 .023 .585 Asian 2.338 .89 .111 1.272 .055 .168 Other 2.835 .073 .217 3.515 .096 .017 Family structure Single parents .804 .050 .428 .068 .055 .916 Step parents .980 .053 .364 .981 .055 .174 Others 2.312 .063 .290 .090 .003 .942 Dating involvement Occasionally .141 .009 .898 1.133 .087 .074 Steady 1.184 .098 .180 .151 .015 .774 O thers 2.307 .098 .131 .103 .005 .912 Campus 1.352 .108 .127 .084 .008 .866 Risky s exual behavior .223 .148 .015 .301 ** .230 .000 Constant 12.38 ** .000 10.8 4** .000 F = 3.791** F = 5.593** p = .000 p = .000 R 2 = 221 R 2 = 164

PAGE 118

118 Table 4 10. OLS regression model for victimization by dating violence (subgroup of victims only) in the U.S. *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating Model A(victims of physical violence) N=218 Model B(victim s of psychological abuse) N=511 B P B P Independent Variables Victim of c hild a buse .297 .109 .152 .095 .036 .458 Witnessed fathers violence .645 .033 .658 .576 .030 .531 Witnessed mothers violence 1.758 .095 .205 .101 .006 .903 Self c ontrol scale .107 ** .188 .007 .056 .111 .014 Attachment to mother .078 .053 .487 .096 .062 .198 Attachment to father .190 .158 .042 .106 .093 .057 Control Variables Sex 1.002 .085 .231 .165 .015 .734 Age .627 .132 .123 .100 .024 .655 Race Hispanic .625 .040 .561 .496 .034 .450 Black 1.371 .083 .242 .210 .014 .768 Asian 1.547 .066 .329 .285 .013 .775 Other 2.277 .067 .352 2.185 .063 .147 Family structure Single parents 1.969 .136 .058 .194 .014 .775 Step parents 2.449 .129 .065 .144 .008 .850 Others 1.551 .045 .534 .012 .000 .993 Dating involvement Occasionally .429 .031 .721 .124 .010 .849 Steady .960 .086 .341 .148 .015 .791 O thers 1.917 .097 .226 .248 .011 .809 Campus 1.333 .110 .204 .382 .036 .485 Risky s exual behavior .263 .183 .011 .330 ** .226 .000 Constant 12.17 ** .000 11.71 ** .000 F = 2.397** F = 3 683** p = .0 1 p = .000 R 2 = 196 R 2 = 129

PAGE 119

119 Table 4 11. Descriptive statistics for full sample Variables Mean Median Mode Sd Min Max Dependent Variable Perpetration of p hysical v iolence 92 .00 .00 3.11 .00 3 6 .00 Perpetration of p sychological v iolence 2.61 .00 .00 4.45 .00 2 8 .00 Victimization by p hysical v iolence .7 2 .00 .00 2. 77 .00 36.00 Victimization by p sychological v iolence 2 49 .00 .00 4.38 .00 2 8 .00 Independent Variables Victim of c hild a buse 1. 39 1.00 .00 1.75 .00 8 .00 Witnessed fathers violence 15 .00 .00 36 .00 1.00 Witnessed mothers violence .08 0 .00 .00 .2 7 .00 1.00 Self c ontrol scale 6 4.91 6 4 .00 63 .00 10.46 23.00 115.00 Attachment to mother 1 7 19 18 .00 20.00 3.3 4 4.00 24.00 Attachment to father 15. 71 1 7 .00 20.00 4. 60 4.00 24.00 Control Variables Sex 60 1.00 1.00 49 .00 1.00 Age 2. 62 3.00 4.00 1.1 8 1.00 4.00 Class attending 2. 20 2.00 1.00 1.0 8 1.00 5.00 Family structure 1. 23 1.00 1.00 61 1.00 4.00 Dating involvement 1. 96 2 .00 1.00 .97 1.00 4.00 Campus .79 1.00 1.00 40 .00 1.00 Risky s exual behavior 3 1 1 2 .00 .00 3. 95 .00 18.00 N= 2736

PAGE 120

120 Table 4 12. OLS regression models for perpetration of dating violence ( full model) in the U.S. and Korea *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Model A(perpetration of physical violence) N= 2456 Model B(perpetration of psychological abuse) N= 2373 B P B P Independent Variables Victim of c hild a buse .192** .106 .000 .288** .112 .000 Witnessed fathers violence .056 .006 .771 .127 .010 .614 Witnessed mothers violence .153 .013 .532 .256 .016 .422 Self c ontrol scale .052** .154 .000 .068 ** .143 .000 Attachment to mother .002 .002 .941 .006 .004 .827 Attachment to father .030 .044 .048 .021 .021 .325 Control Variables Country .459** .073 .004 2.144** .240 .000 Sex .704** .109 .000 1.148** .126 .000 Age .035 .013 .549 .145 .038 .061 Family structure Single parents .138 .012 .541 .135 009 .650 Step parents .652* .041 .037 1.031* .046 .013 Others .504 .021 .281 .735 .022 .233 Dating involvement Occasionally .331 .039 .063 .917** .076 .000 Steady .710** .107 .000 1.518** .163 .000 O thers 1.812** .103 .000 1.948** 4.372 .000 Campus .330 .042 .058 .090 .080 .697 Risky s exual behavior .100** .123 .000 .297** 12.64 4 .000 Constant 3.486** .000 3.228** .000 F = 16.830 F = 47.004 p = .000 p = .000 R 2 = .105 R 2 = 253

PAGE 121

121 Table 4 13. OLS regression models for victimization by dating violence ( full model ) in the U.S. and Korea *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Model A( victimization by physical violence) N= 2463 Model B(victimization by psychological abuse) N= 2384 B P B P Independent Variables Victim of c hild a buse .164** .102 .000 .259** .102 .000 Witnessed fathers violence .102 .013 .557 .094 .008 .708 Witnessed mothers violence .213 .021 .330 .124 .008 .695 Self c ontrol scale .037** .122 .000 .063** .134 .000 Attachment to mother .039* .046 .039 .020 .015 .468 Attachment to father .016 .026 .244 .013 .013 .533 Control Variables Country .377** .068 .008 1.887** .214 .000 Sex .169 .030 .144 .856** .095 .000 Age .019 .008 .719 .100 .027 .194 Family structure Single parents .091 .009 .654 .065 .004 .825 Step parents .156 .011 .574 1.024* .046 .013 Others .159 .007 .706 .664 .020 .279 Dating involvement Occasionally .221 .029 .164 .939** .078 .000 Steady .453** .078 .000 1.328** .144 .000 O thers 1.649** .105 .000 1.952** .081 .000 Campus .132 .019 .395 .061 .006 .791 Risky s exual behavior .085** .118 .000 .320** .282 .000 Constant 3.058 ** .000 3.355 ** .000 F = 12.123 F = 44.300 p = .000 p = .000 R 2 = .078 R 2 = .241

PAGE 122

122 Table 4 14 OLS regression models for perpetration of dating violence in the U.S. and Korea (sub group of perpetrator s only ) *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Model A( perpetration of physical violence) N= 480 Model B( perpetration of psychological abuse) N= 1030 B P B P Independent Variables Victim of c hild a buse .446** .153 .000 .364** .133 .000 Witnessed fathers violence .275 .019 .720 .365 .027 .407 Witnessed mothers violence .914 .049 .316 .207 .012 .683 Self c ontrol scale .144** .250 .000 .103** .199 .000 Attachment to mother .041 .026 .595 .021 .014 .656 Attachment to father .116 .095 .057 .045 .039 .228 Control Variables Country .810 .070 .218 3.209** .317 .000 Sex 1.363* .110 .019 .964** .092 .002 Age .002 .000 .992 .228 .055 .096 Family structure Single parents .167 .009 .843 .272 .016 .589 Step parents .283 .014 .766 .634 .031 .287 Others .040 .001 .981 .965 .027 .340 Dating involvement Occasionally .077 .006 .919 .324 .026 .436 Steady 1.076 .094 .095 .405 .040 .244 O thers 3.735** .151 .002 .999 .041 .169 Campus 1.010 .075 .143 .288 .023 .466 Risky s exual behavior .106 .075 .124 .247** .198 .000 Constant 12.394** .000 8.247** .000 F = 4.475 F = 17.157 p = .000 p = .000 R 2 = .141 R 2 = .224

PAGE 123

123 Table 4 15. OLS regression models for victimization by dating violence in the U.S. and Korea (sub group of victim s only ) *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Model A( victimization by physical violence) N= 376 Model B(victimization by psychological abuse) N= 974 B P B P Independent Variables Victim of c hild a buse .410* .147 .012 .340** .124 .000 Witnessed fathers violence .1.076 .074 .226 .458 .033 .322 Witnessed mothers violence .304 .017 .759 .309 .018 .572 Self c ontrol scale .104** .185 .001 .086** .167 .000 Attachment to mother .016 .011 .852 .057 .039 .243 Attachment to father .136 .112 .061 .066 .056 .092 Control Variables Country .466 .041 .524 2.585** .255 .000 Sex .507 .043 .414 .445 .042 .166 Age .237 .050 .386 .122 .029 .387 Family structure Single parents .1043 .060 .266 .409 .024 .429 Step parents .1016 .046 .387 .461 .023 .446 Others .259 .007 .896 .217 .006 .828 Dating involvement Occasionally .228 .017 .791 .058 .005 .893 Steady .689 .061 .340 .372 .037 .300 O thers 3.442** .150 .008 1.562* .064 .040 Campus 1.492 .106 .073 .455 .037 .270 Risky s exual behavior .076 .054 .338 .285** .230 .000 Constant 12.811** .000 9.712** .000 F = 2.648 F = 14.660 p = .000 p = .000 R 2 = .112 R 2 = .207

PAGE 124

124 T able 4 16. OLS regression models for perpetration of psychological abuse by gender in Korea (o verall m odel) Full Model A(ppsy) N=1278 Male Model B(ppsy) N=577 Female Model C(ppsy) N=713 B P B P B P Independent Variables Victim of c hild a buse .163** .139 .000 .215** .159 .000 .162** .137 .000 Witnessed fathers violence .106 .019 .514 .134 .020 .651 .068 .013 .748 Witnessed mothers violence .055 .007 .816 .448 .045 .303 .244 .030 .431 Self c ontrol scale .041** .142 .000 .022 .072 .085 .061** .187 .000 Attachment to mother .039 .056 .078 .093* .105 .041 .019 .028 .495 Attachment to father .006 .010 .753 .006 .155 .877 .004 .007 .861 Control Variables Sex .361** .077 .006 ----Age .137* .065 .018 .179 .070 .110 .204** 1.00 .005 Family structure Single parents .754* .064 .015 .000 .000 1.00 1.061* .090 .010 Step parents .349 .016 .541 .147 .005 .897 .273 .013 .700 Others .220 .011 .671 .947 .031 .444 .565 .032 .348 Dating involvement Occasionally .811** .124 .000 .011 .031 .976 .975** .158 .000 Steady .774** .153 .000 .260 .044 .322 .823** .165 .000 O thers 1.240** .092 .001 2.867** .203 .000 .216 .014 .693 Campus .492 .011 .092 .740 .070 .086 .235 .016 .636 Risky s exual behavior .114** .171 .000 .021 .029 .507 .162** .198 .000 Constant 3.093** .000 4.101* .000 2.949** .002 F = 13.985 F = 4.437 F = 11.006 p = .000 p = .000 p = .000 R 2 = .151 R 2 = .143 R 2 = .191 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 125

125 Table 4 17. OLS regression models for victimization by psychological abuse by gender in Korea (o verall m odel) *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating Full Model A(vpsy) N=1275 Male Model B(vpsy) N=562 Female Model C(vpsy) N=713 B P B P B P Independent Variables Victim of c hild a buse .199** .110 .000 .304** .176 .000 .123 .066 .083 Witnessed fathers violence .225 .026 .364 .321 .038 .387 .153 .018 .645 Witnessed mothers violence .001 .000 .998 .004 .000 .994 .063 .005 .898 Self c ontrol scale .056** .124 .000 .038* .099 .017 .081** .155 .000 Attachment to mother .011 .010 .742 .072 .066 .199 .025 .024 .559 Attachment to father .034 .040 .211 .003 .003 .957 .048 .057 .165 Control Variables Sex .820** .113 .000 ----Age .265** .082 .003 .256 .078 .069 .317** .097 .005 Family structure Single parents .884 .049 .062 1.372* .079 .049 .730 .039 .260 Step parents 1.290 .093 .138 .136 .004 .922 1.741 .054 .118 Others .143 .005 .856 1.350 .039 .334 .783 .028 .410 Dating involvement Occasionally 1.175** .116 .000 .412 .040 .347 1.539** .155 .000 Steady 1.478** .191 .000 1.377** .183 .000 1.505** .190 .000 O thers .691 .033 .215 1.405 .078 .057 .561 .022 .514 Campus .373** .082 .003 .888 .064 .106 .745 .033 .514 Risky s exual behavior .221** .217 .000 .158** .175 .000 311** .242 .000 Constant 3.588** .000 3.643** .005 4.339** .004 F = 15.536 F=6.701 F=12.166 p = .00 0 p=.000 p =.000 R 2 = .165 R 2 = .155 R 2 = .207

PAGE 126

126 Table 4 18. OLS regression models for victimization by psychological abuse by gender in Korea (subgroup of victims only) Full Model A(vpsy) N=419 Male Model B(vpsy) N=178 Female Model C(vpsy) N=241 B P B P B P Independent Variables Victim of c hild a buse .208* .102 .045 .453** .226 .004 .010 005 .947 Witnessed fathers violence .693 .068 .191 1.508 .146 .068 .070 .007 .947 Witnessed mothers violence .257 .018 .728 .165 .011 .885 .250 .018 .806 Self c ontrol scale .061* .113 .022 .025 .051 .507 .113** .185 .005 Attachment to mother .016 .012 .820 .026 .020 .826 .002 .002 .980 Attachment to father .134* .124 .024 .121 .103 .265 .133 .128 .070 Control Variables Sex .992* .111 .030 ----Age .317 .079 .106 .520 .121 .115 .249 .064 .313 Family structure Single parents 1.618 .076 .108 2.482 .118 .108 .782 .037 .562 Step parents 1.387 .046 .335 .838 .027 .713 3.461 .116 .069 Others .420 .010 .825 .904 .025 .735 .780 .017 .788 Dating involvement Occasionally .538 .049 .364 .704 .058 .485 .846 .082 .258 Steady .971* .110 .047 1.152 .125 .146 .644 .075 .307 O thers 3.113* .097 .046 4.298* .168 .030 .617 .013 .835 Campus 1.463 .067 .160 .1824 .113 .123 .976 .021 .737 Risky s exual behavior .250** .215 .000 .197* .171 .030 .277** .214 .001 Constant 6.079* .039 11.35** .005 F = 4.342 F =2.722 F =2.920 p = .00 p =.001 p =.00 R 2 = .147 R 2 = .201 R 2 = .163 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 127

127 Table 4 19. OLS regression models for perpetration of physical violence by gender in the U.S. (overall model) Full Model A(ppv) N=1158 Male Model B( ppv ) N=393 Female Model B(ppv ) N=765 B P B P B P Independent Variables Victim of c hild a buse .208** .101 .001 .077 061 .247 .262** .110 .005 Witnessed fathers violence .001 .000 .998 1.496** .147 .007 .649 .038 .332 Witnessed mothers violence .404 .029 .367 .265 .033 .528 .945 .053 .167 Self c ontrol scale .068** .172 .000 .048** .086 .000 .075** .163 .000 Attachment to mother .016 .013 .655 .040 .052 .309 .063 .048 .194 Attachment to father .036 .048 .125 .020 .041 .438 .054 .063 .105 Control Variables Sex 1.258** .160 .000 ---Age .053 .016 .618 .002 .001 .986 .103 .028 .485 Race Hispanic .067 .006 .827 .013 .002 .969 .183 .015 .664 Black .742* .066 .029 .163 .020 .683 1.257** .103 .006 Asian .459 .027 .345 .657 .059 .207 .369 .019 .586 Other .912 .040 .155 1.954** .154 .001 .124 .004 .899 Family structure Single parents .018 .002 .955 .003 .000 .993 .096 .008 .832 Step parents .531 .038 .186 .414 .040 .391 .828 .055 .118 Others .163 .006 .822 3.413** .185 .000 .1.582 .057 .103 Dating involvement Occasionally .254 .026 .406 .085 .014 .786 .369 .033 .402 Steady .920** .121 .000 .482 .093 .070 1.089** .128 .002 O thers 1.692** .089 .003 2.439** .198 .000 1.095 .051 .159 Campus .368 .048 .133 .136 .027 .625 .655* .076 .050 Risky s exual behavior .124** .133 .000 .051 .091 .074 .168** .171 .000 Constant 4.060** .000 3.140** .002 4.941** .001 F = 8.716** F =6.328 F =6.573 p = .000 p = .000 p =.000 R 2 = 133 R 2 = 244 R 2 =.144 *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating

PAGE 128

128 Table 4 20. OLS regression models for perpetration of psychological abuse by gender in the U.S. (overall model) Full Model A(ppsy) N=1111 Male Model B(ppsy) N=371 Female Model B(ppsy) N=740 B P B P B P Independent Variables Victim of c hild a buse .389 ** .132 .000 .425** .167 .002 .344** .109 .003 Witnessed fathers violence .407 .019 .534 .235 .012 .841 .651 .030 .411 Witnessed mothers violence .568 .029 .355 .360 .023 .684 1.291 .058 .111 Self c ontrol scale .086 ** .155 .000 .078** .155 .002 .088** .149 .000 Attachment to mother .045 .027 .355 .079 .052 .345 .128* .075 .033 Attachment to father .034 .030 .326 .008 .007 .907 .037 .033 .372 Control Variables Sex 1.636 ** .148 .000 ----Age .083 .018 .563 .095 .023 .686 .076 .016 .677 Race Hispanic .580 .039 .162 .231 .017 .745 .674 .044 .188 Black .529 .033 .248 .1893* .123 .021 .319 .020 .569 Asian .992 .042 .127 .1358 .064 .198 .875 .035 .288 Other .765 .024 .375 .283 .012 .815 1.264 .034 .298 Family structure Single parents .577 .038 .191 .177 .013 .806 .781 .051 .161 Step parents .951 .047 .084 .483 .024 .621 1.465* .073 .028 Others .392 .011 .690 2.166 .057 .250 .147 .004 .899 Dating involvement Occasionally .670 .048 .108 .497 .041 .444 .710 .049 .191 Steady 1.781 .167 .000 .855 .084 .120 2.260** .207 .000 O thers 2.573 ** .099 .000 1.335 .058 .260 3.164** .116 .001 Campus .256 .024 .441 .674 .067 .237 .076 .007 .853 Risky s exual behavior .382 ** .294 .000 .328** .301 .000 .443** .312 .000 Constant 5.146 ** .000 6.846 ** .001 5.186** .004 F = 16.549 F =5.368 F =12.992 p = .000 p =.000 p =.000 R 2 = 233 R 2 =.225 R 2 =.255 *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating

PAGE 129

129 Table 4 21. OLS regression models for victimization by psychological abuse by gender in the U.S. (overall model) Full Model A(vpsy) N=1111 Male Model B(vpsy) N=376 Female Model B(vpsy) N=742 B P B P B P Independent Variables Victim of c hild a buse .296** .111 .000 .491** .191 .000 .173 .064 .089 Witnessed fathers violence .219 .011 .723 1.023 .048 .384 .572 .030 .434 Witnessed mothers violence .250 .014 .654 .134 .009 .878 .372 .019 .616 Self c ontrol scale .056** .113 .000 .058* .117 .018 .056** .109 .002 Attachment to mother .015 .010 .736 .069 .045 .409 .065 .044 .223 Attachment to father .022 .021 .491 .027 .024 .675 .028 .029 .441 Control Variables Sex .720* .072 .011 ------Age .050 .012 .712 .119 .029 .614 .016 .004 .922 Race Hispanic .297 .022 .440 .295 .021 .676 .250 .019 .587 Black .590 .041 .160 2.052** .133 .010 .295 022 .554 Asian .341 .015 .585 1.569 .071 .145 .351 .016 .648 Other .814 .027 .331 .033 .001 .979 1.427 1.125 .205 Family structure Single parents .370 .027 .363 .474 .034 .506 .278 .021 .578 Step parents .460 .026 .361 .938 .047 .325 .946 .055 .110 Others .700 .021 .453 1.445 .037 .441 .485 .015 .652 Dating involvement Occasionally .915* .073 .018 .337 .028 .600 1.220* .096 .012 Steady 1.089** .112 .000 .366 .036 .500 1.549** .163 .000 O thers 1.327 .056 .052 1.134 .047 .347 1.526 .065 .067 Campus .017 .002 .955 .704 .069 .217 .433 .045 .242 Risky s exual behavior .383 .326 .000 .369** .339 .000 .408** .330 .000 Constant 3.871** .002 4.497* .031 3.727* .019 F = 13.329 F =5.865 F =9.763 p = .000 p =.000 p =.0000 R 2 = 195 R 2 =.238 R 2 =.204 *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating

PAGE 130

130 Table 4 22. OLS regression models for perpetration of physical violence by gender in the U.S. (subgroup of perpetrators only) Full Model A(ppv) N=2 88 Male Model B(ppv) N=65 Female Model B(ppv) N=223 B P B P B P Independent Variables Victim of c hild a buse .360 .118 .062 .002 .001 .995 .523* .160 .024 Witnessed fathers violence .532 .025 .394 .419 .026 .824 1.706 .075 .295 Witnessed mothers violence .843 .039 .542 .206 .015 .893 1.980 .078 .270 Self c ontrol scale .174** .276 .000 .039 .072 .493 .154** .233 .001 Attachment to mother .041 .025 .683 .253 .181 .147 .081 .046 .496 Attachment to father .147 .115 .074 .001 .000 .997 .130 .102 .166 Control Variables Sex 2.940** .205 .001 ----Age .599 .117 .101 .074 .016 .883 1.035* .196 .019 Race Hispanic .759 .043 .457 1.347 .096 .328 .782 .042 .525 Black 2.160* .128 .039 1.015 .054 .600 2.509* .152 .033 Asian 2.338 .89 .111 3.636* .173 .049 1.857 .068 .300 Other 2.835 .073 .217 .443 .023 .835 .750 .011 .856 Family structure Single parents .804 .050 .428 .999 .068 .472 1.212 .074 .317 Step parents .980 .053 .364 .838 .035 .701 .772 .043 .524 Others 2.312 .063 .290 35.02** .855 .000 5.650* .159 .020 Dating involvement Occasionally .141 .009 .898 1.020 .083 .435 .189 .012 .891 Steady 1.184 .098 .180 1.058 .104 .328 .759 .061 .484 O thers 2.307 .098 .131 2.192 .126 .321 1.145 .045 .541 Campus 1.352 .108 .127 .854 .070 .528 2.183* .173 .038 Risky s exual behavior .223* .148 .015 .113 .094 .376 .277* .175 .013 Constant 12.38** .000 .176 .971 12.689** .001 F = 3.791 F =6.312 F =3.024 p = .000 p =.000 p =.000 R 2 = 221 R 2 =.727 R 2 =.221 *p<.05. **p< .01. reference group: White reference group: two biological parents reference group: currently not dating

PAGE 131

131 Table 4 23. OLS regression models for perpetration of psychological abuse by gender in the U.S. (subgroup of perpetrators only) Full Model A(p psy ) N=591 Male Model B(pp spy ) N=170 Female Model B(p psy ) N=421 B P B P B P Independent Variables Victim of c hild a buse .446** .157 .000 .593** .234 .006 .365* .122 .020 Witnessed fathers violence .662 .034 .439 .573 .029 .735 .683 .035 .499 Witnessed mothers violence .050 .003 .949 .930 .062 .483 .373 .019 .703 Self c ontrol scale .107** .200 .000 .133** .258 .001 .097** .174 .000 Attachment to mother .056 .036 .399 .096 .059 .505 .117 .076 .129 Attachment to father .046 .040 .367 .014 .013 .892 .058 .049 .350 Control Variables Sex 1.393** .122 .003 ----Age .144 .033 .485 .067 .015 .856 .250 .056 .324 Race Hispanic .228 .015 .711 .274 .019 .811 .413 .027 .580 Black .356 .023 .585 2.134 .128 .122 .371 .024 .628 Asian 1.272 .055 .168 2.122 .090 .243 1.010 .044 .358 Other 3.515* .096 .017 1.654 .061 .459 4.590* .104 .026 Family structure Single parents .068 .055 .916 1.311 .088 .284 .484 .032 .536 Step parents .981 .055 .174 .468 .025 .756 1.578 .090 .063 Others .090 .003 .942 .287 .010 .899 .031 .001 .984 Dating involvement Occasionally 1.133 .087 .074 1.035 .088 .325 1.258 .093 .123 Steady .151 .015 .774 1.028 .102 .263 .323 .031 .627 O thers .103 .005 .912 .330 .015 .845 .100 .005 .929 Campus .084 .008 .866 .267 .023 .779 .219 .020 .715 Risky s exual behavior .301** .230 .000 .296** .256 .002 .324** .231 .000 Constant 10.84** .000 15.45** .000 10.20** .000 F = 5.593 F = 2.668 F = 4.062 p = .000 p =.000 p =.000 R 2 = 164 R 2 =.253 R 2 =.161 *p<.05. **p<.01. reference group: White reference group: two biological parents reference group: currently not dating

PAGE 132

132 Table 4 24. Significance of d ifference in OLS regression coefficients for male and female models in Korea Korea Variables Male (b1) Male (SEb1) Female (b2) Female ( SEb2) Z score Dependent Variables Independent and Control Variables Perpetration of psychological abuse Child abuse 215 .0 59 162 .0 45 714 Victimization by psychological abuse Child abuse .304 .074 .123 .071 1.765* Self control .0 38 .0 16 .0 81 .0 18 1.785* Dating steady 1.377 33 1.505 291 291 Risky sexual behavior 158 .0 39 311 .0 46 2.537** Victimization by psychological abuse (victim group) Risky sexual behavior 197 .0 9 277 .0 85 646 *90% (z=1.65) **95% (z=1.96) ***99%(z=2.58)

PAGE 133

133 Table 4 25. Significance of d ifference in OLS regression coefficients for male and female models in the U.S. U.S. Variables Male (b1) Male (SEb1) Female (b2) Female (SEb2) Z score Dependent Variables Independent and Control Variables Perpetration of physical violence Self control .0 48 .0 12 .0 75 .0 16 1.35 Perpetration of psychological abuse Child abuse .425 .139 .344 .114 .451 Self control .0 78 .0 25 .0 88 .0 2 312 Risky sexual behavior 328 .0 57 443 .0 53 1.478 Victimization by psychological abuse Self control .058 .024 .056 .018 .067 Risky sexual behavior 369 .0 56 408 .0 47 533 Perpetration of physical violence (perpetrator group) Family (other ) 35.021 5.306 5.65 2.402 6.983*** Perpetration of psychological abuse Child abuse .593 .212 .365 .156 .866 (perpetrator group ) Self control 133 .0 39 .0 97 .0 27 759 Risky sexual behavior 296 .0 92 324 .0 7 242 *90% (z=1.65) **95% (z=1.96) ***99%(z=2.58)

PAGE 134

134 Table 4 26. Significance of difference in OLS regression coefficients for Korean and American models Variables KR(b1) KR(SEb1) US(b2) US(SE b2) Z score Dependent Variables Independent and Control Variables Perpetration of physical violence Child abuse .135 .037 .208 .101 .679 Self control .0 37 .0 09 .0 68 .0 12 2.067** Dating steady 524 1 00 92 247 1.486 Dating other 1.851 125 1.692 558 278 Risky sexual behavior .0 56 .0 79 124 .0 29 808 Perpetration of psychological abuse Child abuse .163 .034 .389 .088 2.396** Self control .0 41 .0 08 .0 86 .0 16 2.516** Sex 361 131 1.636 306 3.830*** Dating steady 774 144 1.781 167 4.567*** Dating other 1.24 358 2.573 .0 99 3.589*** Risky sexual behavior 114 .0 19 382 .0 39 6.178*** Victimization by physical violence Child abuse .104 .037 .188 .054 1.283 Self control .0 33 .0 09 .0 41 .0 19 381 Dating steady 455 159 47 207 .0 57 Dating other 1.556 401 1.651 481 152 Victimization by psychological abuse C hild abuse .199 .051 .296 .036 1.554 Self control .0 56 .0 12 .0 56 .0 14 .0 00 Sex 82 2 72 284 288 Dating occasionally 1.175 276 915 385 549 Dating other 1.478 218 1.089 31 1.026 Risky sexual behavior 221 .0 29 383 .0 36 3.504*** Perpetration of physical violence (perpetrator group ) Self control .108 .044 .174 .037 1.148 Perpetration of psychological abuse (perpetrator group) Child abuse .199 .072 .446 .124 1.723* Self control .0 79 .0 18 107 .0 22 985 Victimization by psychological abuse (victim group) Self control .061 .026 .056 .023 .144 Risky sexual behavior 25 .0 6 33 .0 57 967 *90% (z=1.65) **95% (z=1.96) ***99 %( z=2.58)

PAGE 135

135 CHAPTER 5 CONCLUSION Summary Discussion The purpose of this study was to investigate theoretically relevant variables that are linked in some way to the respondents family of origin that may account for the likelihood of both perpetration of and victimization by dating violence in two different cultures. In using the same survey instrument for two countries, this study proposed to test whether theoretical explanations developed in Western society could be applied to Eastern society. A childhood experience of family violence is expected to be positively related to both committing violence a nd being the victim of violence in a dating relationship for both male and female respondents in both in the U.S. and Korea. Parental attachment and self -control (assumed by the theory to be formed in the family of origin at an early age and to be relative ly stable throughout the life course ) are hypothesized to be negatively related to dating violence, again for both perpetrators and victims, and male and female respondents in both countries. The major theory examined is self -control theory which would pre dict that persons with low self -control are more likely to engage in violence against another person and more likely to be a victim of violence by another person The research was designed to test this theory and the concept of self -control was measured in a fairly complete and thorough way. Therefore, this is the only theoretical perspective about which firm conclusions can be reached from the findings. However, the variables of direct (being the victim of violence) and vicarious experience of violence (wi tness violence committed by one family member against another) in the family of origin were seen as consistent with social learning theory and the variable of parental attachment (how close are the relationship with ones parents) was seen as consistent wi th or derivable from social bonding theory. Therefore, the findings were seen as somewhat relevant to these two additional theories and some conclusions

PAGE 136

136 can be offered regarding how well the prior experience with violence and attachment to parents account for dating violence. The results of relating the empirical measures of the explanatory variables to the measures of perpetration and victimization of dating violence will be summarized for the combined samples from the US and Korea, by each country, and by gender. Similarities and differences will be discussed. Self -C ontrol Theory Th e finding that when the different dimensions of self -control were measured separately only the dimension of impulsivity has a significant effect on dating violence is partiall y consistent with Kerly et al. (2008). W hen the different dimensions are placed into a single scale that scale had significant main and net effects on both for both the Korea n and American samples and for the combined samples That is, those students who had low self -control were more likely to be both perpetrators and victims of both physical and psychological dating violence. Significant predictive power was found in most models except for the subgroup of victims in Korea. In particular, the victim s of p hysical violence in Korea had no lower levels of self -control than those who had not been victims However, all models of American sample showed a significant relationship between low self -control and perpetration of and victimization by dating violence. A dditional comparisons for gender difference in the U.S. and Korea were run to find the different effect in each gender group. However, the results failed to report a significant difference. That is, the strong likelihood of low self -control to predict perpetration of and victimization by dating violence has been shown similarly not only among Korean female and male students but among U.S. female and male students as well. Another main purpose of this dissertation was to test whether criminological theories developed in Western societies can explain the dating violence in an Eastern society. These measures of low self -control ha ve an effect in both societies and therefore have some

PAGE 137

137 generalizabililty f rom American to Korean society. In compari ng the effect co efficients for the Korean and American samples the power of self -control to predict dating violence is found to be almost same in both societies However, the significant difference between the magnitudes of the coefficients in the two samples indicates that there are some societal or cultural moderation or conditioning of the effects that low self -control has on current dating violence among univer sity students. That is, although self -control is consistently related to perpetration of physical violence and perpetration of psychological abuse in both countries, the effects are stronger for U.S. students perpetration of dating violence than for Korea n students but not for victimization of dating violence1. Direct and Vicarious Experience with Violence in the Family2 Having witnessed fathers violence against mother, and witness of mothers violence against father was a measure of prior exposure or indirect experience with violence. Having been abused as child was a measure of direct prior experience with violence. Overall, direct experience of child abuse showed a very significant relationship with both perpetration of and victimization by dating violence in both Korea and the U.S. Indirect or vicarious experience of inter -parental violence did not have significant effects on dating violence. The significant predictive power was found in most models except for the subgroup of those who had perpetr ated violence in a dating situation in the past year (versus the subgroup who reported no such behavior in that time period) and the subgroup of only those who had been victims of such violence in the past year In particular, the relationship was not fou nd among the perpetrator s and victim s of physical violence sub groups in Korea. In the U.S., prior experience with violence 1 T test reports that s core of self control of American students are significantly higher than Korean students 2 Although I refer to this exposure to violence in the home as reflective of social learning concepts it is proper to refer to it only intergenerational transmission theory

PAGE 138

138 in the family failed to predict the relationships between physical and psychological abuse in the sub group of victims of physical violence As mentioned, the effect of direct experience of child abuse is significantly related to dating violence in both countries. Additional comparison of coefficients helped to investigate how differently this variable assumed to be consistent with socia l learning theory explains dating violence. Comparison of results of z -scores between the two countries showed that direct experience of child abuse has a more powerful effect on perpetration of psychological abuse among Korean college students compared to U.S. college students. T h is result is repeated for perpetrator group of psychological abuse. Direct experience of child abuse differently explains the perpetration of psychological abuse in the subgroup of perpetrators in Korea and in the U.S. It means th at psychological abuse in dating is explained much more by victimization of child abuse among Korean college students than among American college students In addition child abuse showed the different power by gender to predict dating violence in compari son of coefficient s in Korea. I n particular, g ender effect s in Korea were found in victimization by psychological abuse for all students and in victimization by psychological abuse for the sub group of psychological victims. Direct experience of child abuse has a fairly strong correlation to victimization by psychological abuse for Korean male students compared to Korean female students. Gender effect in the U.S was found in both perpetration of physical and psychological violence for all students and for the subsample of perpetrators. However, there is no significant difference between the two gender groups the effects of direct experience of child abuse in the U.S. The stronger effect of child abuse on dating violence in college is probably related to the higher frequency of child abuse among Korean students They have be en educated to accept the

PAGE 139

139 physical punishment as a form of discipline since childhood. M oreover, they have little chance to b e educa ted how serious psychological abuse can be in dating relationship s There is a n old Korean saying l ittle strokes fell great oaks T his saying is commonly cited for dating relationship. It makes people recognize the abusive relationship as express ion of love and passion. T herefore many people do not identify their wrong relationship as dating violence. It is naturally related to need of education or program of dating violence for college students in Korea. And I hope this dissertation could give an impetus to set up the future education and policy of dating violence. Nevertheless, t his measure of prior experience as the victim of violence in the home has an effect in both societies and therefore has shows some generalizabililty of findings from Ame rican to Korean society. However, the significant difference between the magnitude of the coefficients in the two samples indicates that there some societal or cultural moderation or conditioning of the effects such prior experience has on current dating v iolence among university students. The effect of observing inter -parental violence as a child showed influence only for American male perpetration of physical violence. Specifically, watching a fathers violence against the mother affected perpetration of physical violence only for American male students. Although this is the only finding of gender -specific behavior, it is consistent with previous research of OKeefe (1998). T his indicates that observation of violence between parents has differential effects on violent outcomes later in dating relationships among young adults. This would suggest that there are limitations on the cross -cultural applicability of this as an explanation of dating violence. We have seen its effect in previous research in Am erican samples and we see it again in this study. But it does not apply in Korean sample in this study.

PAGE 140

140 Attachment to Parents Attachment to parents failed to show a significant effect on most models of perpetration of and victimization by dating violence. This was true in cases of physical as well as psychological abuse. Although the effect was not found frequently as often as social learning and self -control variables some model reported the effect of attachment to parents. Interestingly, the effect of la ck of attachment to father was applied for victimization model of victim groups in both countries. N egative effect of attachment to father was observed in case of entire Korean victimization by psychological abuse for victim group of psychological abuse. S imilarly, the negative effect of attachment to father was found in the U.S. among victims group by physical violence and psychological abuse. The effect of attachment to mother was applied for gender model. Although the statistical power is weak relatively low attachment to mother showed the relation to perpetration of psychological abuse for Korean male students. The effect of attachment to mother was significant for U.S. female student to predict a perpetration of psychological abuse. However, it the dir ection was reversed. That is, higher score of attachment to mother was related to perpetration of psychological abuse for U.S. female students. Control Variables In addition to the theoretical or explanatory variables, the statistical models included con trol variables of gender, age, family structure, dating involvement, living campus and risky sexual behavior. Although previous research ha s reported that females were more likely to be victimized in dating partner, female students reported they have more exposure to dating violence as both perpetrators and victims in this research. This is consistent with Gover et al. (2008) finding. Gover et al (2008) interpreted this result as a change of pattern of dating. Other plausible explanation is that females in itiation of violence more related to proximal variables,

PAGE 141

141 such as self defense, as opposed to historical variables, such as child abuse, or inter -parental violence. Some research provided support for this contention, demonstrating that the receipt of violen ce from a partner was a significant predictor of females dating violence (Bookwala et al., 1992; OKeefe, 1997). Although age effect was not found as often as gender effect, the pattern that younger students are more involved in perpetration of physical violence, and older students are involved in perpetration and victimization by psychological abuse was found. Older females reported more experiences of victimization by dating violence. The reason older students might be related to more change to have a dating partner and to have longer period of dating than younger students. In terms of number of dating partner and dating duration, it is possible that this variable has a problem of multicollinearity with a variable of risky sexual behavior because risky s exual behavior is asking the number of sexual partner so far. Students engaged in risky sexual behavior are consistently more likely to experience dating violence as both perpetrator and victim, both in Korea and the U.S and for both female and male. It i s very consistent with previous research showing that extensive dating increase s the likelihood of having been in an abusive relationship (Bergman, 1992; Reuterman & Burcky, 1989; Sugarm a n & Hoteling, 1989). It is unclear, however, whether extensive dating is a risk because frequently dating students extensive dating experience simply increase their probability of pa i ring with at least one abusive partner among their many partners or whether the extensive dating histories of some individuals are proxies fo r other factors that increase the likelihood of abuse in relationship (Cleaveland et al. 2003; Lewis & Fremouw, 2001; Rowe et al., 2997) Although there are previous researches that have indicated the effect of divorced parents and single parents on dating violence (Billingham & Notebart,1993; Gover, 2004), m ost models did

PAGE 142

142 not indicate a relationship between family structures growing up and dating violence, with these exceptions: the Korean male group and the Korean victim groups showed the relationsh ip between perpetration of psychological abuse and family structure. Those students who were raised by single parents were more likely to become involved in perpetration of psychological abuse than were other victims. However one should be careful when ex amining the effect of being raised by single parents because over the 95% of Korean students answered that they were raised by in a family with both biological parents. Also, U.S. female students showed the link between perpetration of psychological abuse and stepparents Those female students who were raised by stepparents were more likely to become involved in perpetration of psychological abuse. Although the effects of family structure were not found as frequently as other independent variables, it did seem to affect the incidence of psychological abuse against their partners. Being raised by a single parent was a significant predictive factor for Korean students3, as was being raised by stepparents for U.S. students. This family effect is found on full model including U.S. and Korea sample together. Although the effect of family structure in which students were raised is not found frequently in separate model, the full model shows the effect of family structure. Especially being raised under step parents has significant effect on perpetration and victimization by dating violence in the full model. Dating relationship factors mean a number of dating partner, seriousness of dating relationship, period of dating relationship, and empirical researches have re ported the positive relationship between those dating relationship factors and perpetration of dating violence. Moreover, the association is sometimes functioning differently by gender. Consistently to 3 Most of Korean students (95%) responded that they were raised by both biological parents. It may reflect the low divorce rate at 1980s and early 1990s in Korea. However, the number of who are raised by single parents, stepparents or other relatives are relatively very small, this question may not a good measure for Korean sample.

PAGE 143

143 previous research that reported e xtensive dating incre ases the likelihood of having been in an abusive relationship as well (Bergman, 1992; Reuterman & Burcky, 1989; Sugarman & Hoteling, 1989) this dissertation found that having a serious dating partner has a positive effect on perpetration and victimization by dating violence. However it is dangerous to say that having a serious dating partner increased the likelihood of perpetration and victimization by dating violence based on this dissertation. A s stated above, having a dating partner is a necessary condition for dating violence; used control variable was current dating status and dependent variable was past experience of dating violence. I cannot say positively for certain the unclear relatio nship if students who answered not currently dating were significantly involved in dating violence as a perpetrator or a victim in the past 2) used reference group was currently not dating. Korean results show that that living on campus actually incre ased the likelihood of perpetration of and victimization by dating violence. Although the reasons will need to be investigated in the future, two possible explanations exist for the high numbers of perpetration and victimization among students living on ca mpus in Korea. The first reason for these results is the location of Kyonggi University. Because Kyonggi University is located in a metropolitan area, many students might commute from their own homes and continue to live with their parents. Only students w ho came from out of town might be eligible to be admitted to residence halls. And living on -campus increased the risky campus life than commuting from own house living with family. Second reason is small number of students belonged to living on-campus in Korean sample. Due to the limited number of students admitted to residence halls, the s ample size for living on -campus, especially in subsample models, might be invalid for comparison.

PAGE 144

144 The both data set of Korean and American samples shows some missing v ariables. A lthough the missing variables a re less than 10% of overall data the reason should be considered for future research. although the reason of missing variables are not clear, the reason of missing might be related to sensitive questions such as e xperience of child abuse, sexual behavior and/or other perpetration or victimization of sexual abuse on dating violence In addition, I found the number of missing variables increase at the end of questions of survey instruments and at the questions having long items such as dating violence. Therefore another possible reason of missing variables is from fatigue of long questions. Limitations Although this study contributes to the study of dating violence in both countries, and to knowledge of what f actors account for this behavior across two different societies, its limitations could not be avoided. T he data collected in the project do not allow measurement of all relevant dimensions of social learning theory such as differential association, definition, differential reinforcement and imitation. Early childhood i mitation or violence was measured indirectly by asking about observation of intra -family violence and to a less extent being the victim of intra family violence. M oreover, while the dimensions of Akers s social learning theory relies on both past and present operation of the learning process to account for current behavior the only measurement s in this study were of past events as predictors of current deviant behavior. A lso, the full examina tion of social bonding may require reflecting all dimensions such as attachment, commitment involvement and belief M oreover, attachment dimension may need to include several close relationships such as family, friend, or peers. H owever, this study had not only had only attachment dimension, but also measured one dimension of attachment namely to parents T he result of factor analysis for self -control showed good reliability in this study. H owever, Higgins (2007) did not find unidimensionality of the Gr ansmick et al. (1993) scale with 4 point

PAGE 145

145 Likert -type response categories T hat is, although the Grasmick et al ( 1993) scale of self -control showed reliability, validity is still open to question. Therefore the findings cannot be co nsidered adequate tests of all three theories. The findings here are at best partial and suggestive with regard to how well these theoretical perspectives can account for dating violence. Second, the sample that I used cannot be considered to be represent ative because it was conducted among college students in only one university in Korea and just one university in the U.S. In addition, the sampling method was convenience sampling, hence it is not random sampling. Third, the main dependent variables were m easured using the Revised Conflict Tactic Scale, which, although the most commonly used to measure domestic violence, does not reflect expanded definitions of dating violence s uch as digital dating violence. Moreover, although the items on experience of v iolence in childhood refer to an earlier time period and supposedly self -control was formed in childhood prior to the current time the research is cross -sectional, and this places limitations on interpretations of casual order. In addition, as previously mentioned, the items measuring dating violence asked about such experiences in the past year while the items on being in a dating relationship and how much one is involved in such relationship asked only about current relationships. The proper measure w ould be to make the time frame items measuring victimization by and perpetration of dating violence the same as the time frame for items measuring not being in a dating relationship, dating to some extent, and being more extensively involved in dating. P olicy implication Overall findings of this study emphasize the role of family experience in childhood to predict future dating violence. All of main independent variables such as direct experience of child abuse, low self -control and attachment to parents is formed in the family in childhood. Especially

PAGE 146

146 the most significant factor in this study was self -control, and this self -control is set up by proper parenting in childhood. Low self -control is formed, according to the theory in childhood, and remains s table thereafter. Therefore, o nly policies that take effect early in life and have a positive impact on families have much chance of affecting self -control reducing crime and delinquency (Akers & Sellers, 2004). H owever, there are some possible suggest ions for programs and policies based on the findings of this study that go beyond early childhood intervention. O ne of finding indicates that experience of dating violence among college students was high that cannot be ignored on the campus. M oreover, fem ale students perpetration is higher than male students in the U.S. Although many universities have provided the education program or victim service for college students the education or the program may need to be designed to encourage the participants of male students For the Korean situation, this study may be a pioneering study to examine Korean dating violence among college students. Due to lack of awareness of seriousness of dating violence in Korea there is probably no university which has provided an education or prevention program for college students regarding dating violence The findings in this research showing both male and female involvement and some similarity in predictive variables in both countries I would like to suggest develo ping education program s on campus in Korea n university that borrow from programs that may already be in place in the United States and that programs in each country reflect knowledge of the predictive and causative factors in the behavior Future Researc h Future research on dating violence could take many directions. A lthough I included the variable of risky sexual behavior as a control variable in this study, it is possible the inclusion of this variable unduly weaken s the net effect of other independent variables because it still retains some vulnerability to measurement tautology. I suggest testing the power of independent

PAGE 147

147 variables without risky sexual behavior for future research. Although the current dissertation can contribute to b uilding a theoretical background of dating violence, it did not reflect theoretical perspective s very well beyond self -control theory. Future research could develop good measures of social learning and bonding theory as well as measures of strain, social disorganization, feminist, and other theories that have potential to shed light on accounting for variations in dating violence and provide tests of these criminological theories as explanations of dating violence. It will also be important to examine datin g violence experienced through electronic media. Many students have shared their pictures and routines with their students and dating partners; however, the research related to cyber -violence such as cy -defamation, either by dating violence or by ex -partne rs, has not been studied. Future research is recommended to include those factors related to digital dating violence. Furthermore, it would be important to develop more adequate scales to reflect cultural difference for dating violence or child abuse and t o more adequately test the cross -cultural validity of different theoretical models of dating violence.

PAGE 148

148 APPENDIX A SURVEY INSTRUMENT DEMOGRAPHICS SECTION A Please answer the following questions about your background. Circle the number that corresponds to the answer that you select. A1 What is your sex? 1 Male 2 Female A2 What is your age? 1 18 years old 2 19 years old 3 20 years old 4 21 or older A3 What is your class standing? 1 Freshman 2 Sophomore 3 Junior 4 Senior 5 Graduate A4 What type of household did you mostly live in while you grow up? 1 I lived with my two biological parents 2 I lived with a single parent 3 I lived with a parent and a stepparent 4 I lived with adoptive parents 5 Ot her relatives (grandparents, aunt/uncle, siblings, etc.) 6 Other A5 Where do you live? 1 on campus 2 off campus A6 What is your current relational status? 1 Not currently Dating 2 Occasionally Dating 3 Steady/Exclusively Dating 4 Engaged 5 Mar ried 6 Divorced 7 Other A7 How often do you see the person you are dating? 1 Not currently dating 2 A few times a year 3 Once or twice a month 4 Once a week 5 Twice a week 6 Three or more times a week A 8 Are you currently living with the person you are dating? 1 Yes 2 No 3 I am not currently dating someone A9 What is your race? 1 White, nonHispanic 2 White, Hispanic

PAGE 149

149 3 Black or African American 4 Asian 5 Pacific Islander 6 American Indian or Alaskan Native 7 Other FAMILY OF ORIGIN VIOLENCE SECTION B Now we are going to ask you some questions about violence you may have experienced as a child. Think back to your childhood for how often each of the following behaviors occurred. If your mother or father did not raise you, please answer the questions about the person who did raise you. When you were a child did any parent, stepparent or guardian ever... Yes No B1 Throw something at you that could hurt you? 1 2 B2 Push, grab, or shove you? 1 2 B3 Pull your hair? 1 2 B4 Slap or hit you? 1 2 B5 Kick or bite you? 1 2 B6 Choke or attempt to drown you? 1 2 B7 Hit you with some object? (not including spanking) 1 2 B8 Beat you up? 1 2 B9 Punish you with a belt, board, cord, or other hard object? (not including spanking) 1 2 B10 Hit you so hard that it left bruises or marks? 1 2 B11 Touched you in a sexual way when you did not want that to happen? 1 2 B12 Hurt you in a sexual way? 1 2 B13 When you were a child, did you ever see your father hit your mother? 1 Yes 2 No B14 When you were a child, did you ever see your mother hit your father? 1 Yes 2 No B15 When you were a child, did any parent, stepparent, or guardian ever spank you as a form of discipline? 1Yes 2 No RELATIONSHIP BEHAVIORS SECTION F No matter how well a couple gets along, there are times when they disagree, get annoyed with the other person, or just have spats or fights because they are in a bad mood. This is a list of things that might happen when you and your partner are not getti ng along. Please circle how many times you did each of these things in the past year and how many time s your partner did them to you in the past year If you or your partner did not do one of these things in the past year, but it happened before that, circle 5. How often did this happen in the past year? 0 = This has never happened 1 = Once in the pas t year 2 = Twice in the past year 3 = 35 times in the past year 4 = 6 or more times in the past year 5 = Not in the past year, but it did happen before

PAGE 150

150 This has never happened Once in the past year Twice in the past year 3 5 times in the past year 6 or more times in the past year Not in the past year, but it did happen before F1 Threw something at my partner that could hurt. 0 1 2 3 4 5 F2 My partner did this to me. 0 1 2 3 4 5 F3 Twisted my partners arm or hair. 0 1 2 3 4 5 F4 My partner did this to me. 0 1 2 3 4 5 F5 I kicked my partner. 0 1 2 3 4 5 F6 My partner did this to me. 0 1 2 3 4 5 F7 I slapped my partner. 0 1 2 3 4 5 F8 My partner did this to me. 0 1 2 3 4 5 F9 I pushed or shoved my partner. 0 1 2 3 4 5 F10 My partner did this to me. 0 1 2 3 4 5 F11 I punched or hit my partner with my hand or an object. 0 1 2 3 4 5 F12 My partner did this to me. 0 1 2 3 4 5 F13 I choked my partner. 0 1 2 3 4 5 F14 My partner did this to me. 0 1 2 3 4 5 F15 I slammed my partner against a wall. 0 1 2 3 4 5 F16 My partner did this to me. 0 1 2 3 4 5 F17 I grabbed my partner. 0 1 2 3 4 5 F18 My partner did this to me. 0 1 2 3 4 5 F19 I threatened to hit or throw something at my partner. 0 1 2 3 4 5 F20 My partner did this to me. 0 1 2 3 4 5 F21 I prevented my partner from seeing family or friends. 0 1 2 3 4 5 F22 My partner did this to me. 0 1 2 3 4 5 F23 I insisted on knowing my partners whereabouts all the time. 0 1 2 3 4 5 F24 My partner did this to me. 0 1 2 3 4 5 F25 I insisted on knowing who my partner was talking to on the phone. 0 1 2 3 4 5 F26 My partner did this to me. 0 1 2 3 4 5 F27 I insulted or swore at my partner. 0 1 2 3 4 5 F28 My partner did this to me. 0 1 2 3 4 5 F29 I accused my partner of being a lousy lover. 0 1 2 3 4 5 F30 My partner did this to me. 0 1 2 3 4 5 F31 I called my partner bad names. 0 1 2 3 4 5 F32 My partner did this to me. 0 1 2 3 4 5 F33 I shouted or yelled at my partner. 0 1 2 3 4 5 F34 My partner did this to me. 0 1 2 3 4 5 F35 I used threats to make my partner have sex. 0 1 2 3 4 5 F36 My partner did this to me. 0 1 2 3 4 5 F37 I made my partner have sex without a condom. 0 1 2 3 4 5 F38 My partner did this to me. 0 1 2 3 4 5 F39 I used force (like hitting, holding down, or using a weapon) to make my partner have sex (vaginal, oral, and/or anal). 0 1 2 3 4 5 F40 My partner did this to me. 0 1 2 3 4 5 F41 I insisted on sex (vaginal, oral, and/or anal) when my partner did not want to (but did not use physical force). 0 1 2 3 4 5 F42 My partner did this to me. 0 1 2 3 4 5

PAGE 151

151 SELF CONTROLSECTION H Strongly Agree Agree Disagree Strongly Disagree H1 I often act on the spur of the moment without stopping to think. 1 2 3 4 H2 I dont devote much thought and effort to preparing for the future. 1 2 3 4 H3 I often do whatever brings me pleasure here and now, even at the cost of some distant goal. 1 2 3 4 H4 Im more concerned with what happens to me in the short run than in the long run. 1 2 3 4 H5 I frequently try to avoid projects that I know will be difficult. 1 2 3 4 H6 When things get complicated, I tend to quit or withdraw. 1 2 3 4 H7 The things in life that are easiest to do bring me most pleasure. 1 2 3 4 H8 I dislike really hard tasks that stretch my abilities to the limit. 1 2 3 4 H9 I like to test myself every now and then by doing something a little risky. 1 2 3 4 H10 Sometimes I will take a risk just for the fun of it. 1 2 3 4 H11 I sometimes find it exciting to do things for which I might get in trouble. 1 2 3 4 H12 Excitement and adventure are more important to me than security. 1 2 3 4 H13 If I had a choice, I would almost always rather do something physical than something mental. 1 2 3 4 H14 I almost always feel better when I am on the move than when I am sitting and thinking. 1 2 3 4 H15 I like to get out and do things more than I like to read or contemplate ideas. 1 2 3 4 H16 I often look out for myself first, even if it means making things difficult for other people 1 2 3 4 H17 Im not very sympathetic to other people when they are having problems. 1 2 3 4 H18 If things I do upset people, its their problem not mine. 1 2 3 4 H19 I will try to get the things I want even when I know it causes problems for other people. 1 2 3 4 H20 I lose my temper pretty easily. 1 2 3 4 H21 Often, when Im angry at people I feel more like hurting them than talking to them about why I am angry. 1 2 3 4 H22 When Im really angry, other people better stay away from me. 1 2 3 4 H23 When I have a serious disagreement with someone, its usually hard for me to talk calmly about it without getting upset. 1 2 3 4

PAGE 152

152 RISK FACTORS SECTION G The next set of questions asks you about your past behavior. Again, please remember that this is a confidential survey and your answers cant be linked to you. Please answer the following questions honestly. G1 How old were you when you had sexual inter course for the first time? 1 Thirteen or younger 2 Fourteen 3 Fifteen 4 Sixteen 5 Seventeen 6 Eighteen or older 7 I have never had sexual intercourse G2 During your life, how many people have you had sexual intercourse with? 1 1 partner 2 2 partners 3 3 p artners 4 4 partners 5 5 partners 6 6 partners or more 7 I have never had sexual intercourse G3 During the last three months, how many people have you had sexual intercourse with? 1 1 partner 2 2 partners 3 3 partners 4 4 partners 5 5 partners 6 6 partners or more 7 I did not have sexual intercourse during the past 3 months 8 I have never had sexual intercourse ATTACHMENT TO PARENTAL FIGURES SECTION I Please answer the following questions about your current relationship with your parents, steppa rents, or guardians. ATTACHMENT TO MOM Please answer these questions about your mom/female guardian Never Sometimes Half of the Time Usually Always I1 How often do you trust your mom/female guardian? 1 2 3 4 5 I2 How often do you feel you can talk to her about your problems? 1 2 3 4 5 I3 How often do you think she is genuinely interested in you? 1 2 3 4 5 I4 How often do you feel that she supports you? 1 2 3 4 5 ATTACHMENT TO DAD Please answer these questions about your dad/male guardian Never Sometimes Half of the Time Usually Always I5 How often do you trust your dad/male guardian? 1 2 3 4 5 I6 How often do you feel you can talk to him about your problems? 1 2 3 4 5 I7 How often do you think he is genuinely interested in you? 1 2 3 4 5 I8 How often do you feel that he supports you ? 1 2 3 4 5

PAGE 153

153 APPENDIX B CODE BOOK 1 ) Categorical data information VARIABLE S Gender Male 0 Female 1 Age 18 1 19 2 20 3 21 4 Family structure Two biological parents Reference group Single parents Step parents other Dating No current dating Reference group Occasional Steady/serious dating Other (engaged, married, divorced) Living on/off campus On campus 0 Off campus 1 Race white Reference group Black Hispanics Asian Others Country Korea 0 America 1 Witnessed fathers violence No Yes 0 1 Witnessed mothers violence No Yes 0 1 2 ) Scale information INDEPENDENT VARIABLES Child abuse (KOREA) No Yes 0 1 1 Throw something at you that could hurt you? 2. Push, grab, or shove you? 3. Pull your hair? 4. Slap or hit you? 5. Kick or bite you? 6. Choke or attempt to drown you? 7. Hit you with some object? (not including spanking) 8. Punish you with a belt, board, cord, or other hard object? (not including spanking) 9. Hit you so hard that it left bruises or marks? Child abuse (USA) No Yes 0 1 1 Throw something at you that could hurt you? 2. Push, grab, or shove you?

PAGE 154

154 3 Pull your hair? 4. Slap or hit you? 5. Hit you with some object? (not including spanking) 6. Punish you with a belt, board, cord, or other hard object? (not including spanking) 7 Hit you so hard that it left bruises or marks? Self control Strongly agree 1 1 I often act on the spur of the moment without stopping to think. 2. I dont devote much thought and effort to preparing for the future. 3. I often do whatever brings me pleasure here and now, even at the cost of some distant goal. 4. Im more concerned with what happens to me in the short run than in the long run. 5. I frequently try to avoid projects that I know will be difficult. 6. When things get complicated, I tend to quit or withdraw. 7. The things in life that are easiest to do bring me most pleasure. 8. I dislike really hard tasks that stretch my abilities to the limit. 9. I like to test myself every now and th en by doing something a little risky. 10. Sometimes I will take a risk just for the fun of it. 11. I sometimes find it exciting to do things for which I might get in trouble. 12. Excitement and adventure are more important to me than security. 13. If I had a choice, I would almost always rather do something physical than something mental. 14. I almost always feel better when I am on the move than when I am sitting and thinking. 15. I like to get out and do things more than I like to read or contemplate i deas. 16. I often look out f or myself first, even if it means making things difficult for other people 17. Im not very sympathetic to other people when they are having problems. 18. If things I do upset people, its their problem not mine. 19. I will try to get the things I want even when I know it causes problems for other people. 20. I lose my temper pretty easily. 21. Often, when Im angry at people I feel more like hurting them than t alking to them about why I am angry. 22. When Im really angry, other people better stay away from me. 23. When I have a serio us disagreement with someone, its usually hard for me to talk calmly about it without getting upset. Agree 2 disagree 3 Strongly disagree 4 Attachment to mother Never 1 1 How often do you trust your mom/female guardian? Sometimes 2

PAGE 155

155 Half of the times 3 2 How often do you feel you can talk to her about your problems? 3. How often do you think she is genuinely interested in you? 4 How often do you feel that she supports you? usually 4 Always 5 Attachment to father Never 1 1 How often do you trust your dad/male guardian? 2. How often do you feel you can talk to him about your problems? 3. How often do you think he is genuinely interested in you? 4. How often do you feel that he supports you ? Sometimes 2 Half of the times 3 usually 4 Always 5 Risky sexual behavior No experience 1 partner (13yrs ) 2 partners (14yrs) 3 partners(15yrs) 4 partners(16yrs) 5 partners (17yrs) 6 partners or more (18 yrs or older) 0 1 2 3 4 5 6 1 How old were you when you had sexual intercourse for the first time? 2 During your life, how many people have you had sexual intercourse with 3 During the last three months, how many people have you had sexual intercourse with DEPENDENT VARIABLES Perpetration of physical violence (KOREA) N ever h appened 0 1 I t hrew something at my partner that could hurt. 2. Twisted my partners arm or hair. 3. I kicked my partner. 4. I slapped my partner. 5. I pushed or shoved my partner. 6. I punched or hit my partner with my hand or an object. 7. I choked my partner. 8. I slammed my partner against a wall. 9. I grabbed my partner. Once in the past yr 1 Twice in the past y r 2 3 5 times in the past yr 3 6 or more times in the past year 4 Perpetration of psychological abuse (KOREA) 1 I threatened to hit or throw something at my partner. 2. I prevented my partner from seeing family or friends. 3. I insulted or swore at my partner. 4. I accused my partner of being a lousy lover. 5. I called my partner bad names. 6. I shouted or yelled at my partner. Victimization of physical violence(KOREA) 1 My partner t hrew something at m e that could hurt. 2. My partner t wisted my arm or hair. 3. My partner kicked m e 4. My partner slapped me 5. My partner pushed or shoved me 6. My partner punched or hit me with my hand or an object. 7. My partner choked me 8. My partner slammed me against a wall. 9. My partner grabbed me

PAGE 156

156 Victimization of psychological violence (KOREA) 1 My partner insisted on knowing me whereabouts all the time. 2. My partner prevented me from seeing family or friends. 3. My partner insisted on knowing who I was talking to on the phone. 4. My partner insulted or swore at me 5. My partner accused m e of being a lousy lover. 6. My partner shouted or yelled at me Perpetration of physical violence (US) 1 I t hrew something at my partner that could hurt. 2. Twisted my partners arm or hair. 3. I kicked my partner. 4. I slapped my partner. 5. I pushed or shoved my partner. 6. I punched or hit my partner with my hand or an object. 7. I choked my partner. 8. I slammed my partner against a wall. 9. I grabbed my partner Perpetration of psychological abuse (US) 1 I threatened to hit or throw something at my partner. 2. I insisted on knowing who my partner was talking to on the phone. 3. I insulted or swore at my partner. 4. I accused my partner of being a lousy lover. 5. I called my partner bad names. 6. I shouted or yelled at my partner. Victimization of physical violence (US) 1 My partner t hrew something at m e that could hurt. 2. My partner t wisted my arm or hair. 3. My partner kicked m e 4. My partner slapped me 5. My partner pushed or shoved me 6. My partner punched or hit me with my hand or an object. 7. My partner choked me 8. My partner slammed me against a wall. 9. My partner grabbed me Victimization of psychological abuse (US) 1 My partner prevented me from seeing family or friends. 2. My partner insisted on knowing me whereabouts all the time. 3. My partner insisted on knowing who I was talking to on the phone. 4. My partner insulted or swore at me 5. My partner accused me of being a lousy lover. 6. My partner called me bad names.

PAGE 157

157 APPENDIX C CORRELATION CHARTS

PAGE 158

158 Correlation c hart for Korean s ample 1 child abuse 2 Dad violence 3 Moms violence 4 self control 5 Att mom 6 Att dad 7 gender 8 age 9 campus 10 risky sex 11 family 12 dating 13 ppv 14 ppsy 15 vpv 16 vpsy 1 1.00 2 .29** 1.00 3 .27** .36** 1.00 4 .073** 0.02 0.02 1.00 5 .18** .08** .073** .22** 1.00 6 .22** .17** .097** .21** .56** 1.00 7 0.02 0.01 0.01 0.01 .06* .15** 1.00 8 0.04 0.00 0.02 .13** .09** .08** .15** 1.00 9 0.01 .08** 0.01 0.01 0.04 .06* .10** 0.00 1.00 10 .10** 0.02 0.03 .07** 0.01 0.02 .31** .26** 0.02 1.00 11 0.04 .10** 0.01 0.05 .09** 0.05 0.02 0.02 0.00 .05* 1.00 12 0.05 0.05 .062* 0.02 0.05 0.01 .07** .16** 0.03 .22** 0.03 1.00 13 .14** .07** .064* .11** .06* .07** 0.00 0.04 0.04 .10** .05* .13** 1.00 14 .18** .07** .070* .12** .09** .07** 0.00 .09** 0.02 .23** 0.05 .21** .60** 1.00 15 .10** 0.04 .055* .07** .05* 0.05 0.04 0.01 .05* .09** 0.02 .12** .87** .54** 1.00 16 .14** 0.02 0.04 .12** .06* .07** 0.02 .12** 0.00 .27** 0.04 .23** .44** .74** .42** 1.00 **Correlation is significant at the 0.01 level (2tailed) *Correlation is significant at the 0. 01 level (2tailed)

PAGE 159

159 Correlation chart for American sample 1 child abuse 2 Dad v iolence 3 Moms violence 4 self control 5 Att mom 6 Att dad 7 gender 8 age 9 race 10 campus 11 risky sex 12 family 13 dating 14 ppv 15 ppsy 16 vpv 17 vpsy 1 1 2 .28** 1.00 3 .25** .39** 1.00 4 -.10** 0.01 0.00 1.00 5 -.22** -0.05 -.09** .24** 1.00 6 -.21** -.17** -.08** .16** .31** 1.00 7 -.07** 0.01 -.07** .10** -0.01 -.09** 1.00 8 .07** -0.03 -0.01 -0.01 -0.02 -0.05 -.05* 1.00 9 .21** .12** .11** .05* -0.04 -.082** -0.02 0.02 1.00 10 0.04 0.00 -0.01 -0.02 0.01 -0.04 -0.02 .48** 0.02 1.00 11 .13** 0.05 .07* -.14** -.06* -.08** -.07** .27** -0.01 .17** 1.00 12 .12** .10** .06* -0.02 -.14** -.18** 0.02 .05* .057* 0.04 .202** 1.00 13 .05* 0.03 0.05 0.05 -0.02 -0.02 .06* .17** -0.02 .10** .293** .057* 1.00 14 .15** 0.04 0.01 -.17** -.07** -.12** .13** 0.05 .072* 0.00 .196** .09** .17** 1.00 15 .17** 0.04 .07* -.17** -.06* -.08** .09** .13** 0.00 .09** .373** .11** .27** .56** 1.00 16 .16** 0.05 .06* -.16** -.12** -.10** 0.0 .08** .058* 0.04 .209** 0.04 .14** .81** .48** 1.00 17 .15** 0.04 .06* -.15** -.07* -.07* 0.0 .11** 0.01 .07** .387** .10** .21** .45** .84** .46** 1.00 **Correlation is significant at the 0.01 level (2tailed) *Correlation is significant at the 0. 01 level (2tailed)

PAGE 160

160 Correlation chart for full model 1 Child abuse 2 Dad's violence 3 Mom's violence 4 Self control 5 Attach mom 6 Attach dad 7 Gender 8 Age 9 Country 10 c ampus 11 Risky sex 12 F amily 13 Dating 14 Enppv 15 Enppsy 16 Envpv 17 Envpsy 1 1.00 2 .27** 1.00 3 .26** .36** 1.00 4 -.07** -.07** -0.02 1.00 5 -.19** -.10** -.08** .27** 1.00 6 -.21** -.18** -.09** .21** .44** 1.00 7 -0.02 -0.03 -.04* .09** -0.02 -.10** 1.00 8 0.01 .049* 0.01 -0.04 -0.01 -0.02 -.13** 1.00 9 0.02 -.22** -0.02 .32** .17** .11** .11** -.27** 1.00 10 0.02 .06** 0.00 -.14** -.07** -.08** -0.03 .38** -.40** 1.00 11 .11** -0.03 .04* -0.02 0.00 -0.01 -.16** .18** .23** 0.00 1.00 12 .09** .040* 0.02 .07** -.07** -.10** .04* -0.02 .21** -.06** .19** 1.00 13 .05** 0.01 .05** .06** 0.03 0.01 0.00 .12** .10** 0.03 .27** .04* 1.00 14 .14** 0.03 0.03 -.10** -.05** -.09** .08** -0.02 .09** -.04* .17** .09** .16** 1.00 15 .16** -0.02 .05** -.05** -0.02 -.05* .08** 0.03 .27** -.05** .37** .14** .26** .56** 1.00 16 .14** 0.02 .05** -.09** -.08** -.07** -0.01 0.03 .05** -0.01 .16** .042* .14** .83** .48** 1.00 17 .15** -0.02 .04* -.06** -0.03 -.044* .047* 0.03 .25** -.05* .38** .14** .25** .51** .90** .50** 1.00 **Correlation is significant at the 0.01 level (2tailed) *Correlation is signifi cant at the 0.01 level (2tailed)

PAGE 161

161 APPENDIX D TABLES OF LOGISTIC REGRESSION MODELS Table D 1 Logistic regression model for perpetration of physical violence in Korea Variables Valid N:1304 B P OR Independent Variables Victim of c hild a buse .128* .002 .1.136 Witnessed fathers violence .224 .277 1.251 Witnessed mothers violence .115 .687 1.122 Self control scale .037** .001 .964 Attachment to mother .007 .818 .993 Attachment to father .024 .278 .976 Control Variables Sex .306 .093 1.358 Age .011 .886 .989 Family structure .506 Single parents .175 .659 1.191 Step parents .676 .260 1.966 Others .628 .316 1.874 Dating involvement .000 Occasionally 1.318** .000 3.738 Steady .927** .000 2.528 O thers .993* .026 2.699 Campus .440 .229 .644 Risky Sexual behavior .087** .000 1.091 Constant .058 .948 1.060 Model Chi Sq 121.520 2LL 971.845 Cox&Snell R Square .089 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 162

162 Table D 2 Logistic regression model for perpetration of psychological abuse in Korea Variables N=1399 Missing:121 B Valid N:1278 P OR Independent Variables Victim of c hild a buse .137** .000 1.147 Witnessed fathers violence .249 .140 1.283 Witnessed mothers violence .245 .140 1.283 Self control scale .023** .008 .977 Attachment to mother .043 .067 .958 Attachment to father .009 .641 1.009 Control Variables Sex .445** .002 1.577 Age .171 ** .007 1.187 Family structure .703 Single parents .231 .474 1.259 Step parents .004 .995 1.004 Others .560 .335 .571 Dating involvement .000 Occasionally .888** .000 2.430 Steady 1.004** .000 2.730 O thers .352 ** .000 1.422 Campus .160 .611 .852 Risky s exual behavior .116** .000 1.123 Constant .449 .519 .638 Model Chi Sq 183.855 2LL 1404.625 Cox&Snell R Square .134 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 163

163 Table D 3 Logistic regression model for victimization by physical violence in Korea Variables N=1399 Missing:91 B Valid N:1308 P OR Independent Variables Victim of c hild a buse .111* .011 1.117 Witnessed fathers violence .128 .571 1.136 Witnessed mothers violence .165 .592 1.179 Self c ontrol scale .051** .000 .950 Attachment to mother .049 .103 .952 Attachment to father .007 .779 .993 Control Variables Sex .536** .007 1.710 Age .010 .905 1.010 Family structure .413 Single parents .101 .827 .904 Step parents .910 .130 2.484 Others 0497 .459 1.644 Dating involvement .000 Occasionally .906** .000 2.474 Steady .865** .000 2.374 O thers .718 .150 2.050 Campus .521 .189 .594 Risky s exual behavior .091** .000 1.095 Constant 10141 .233 3.129 Model Chi Sq 104.150 2LL 859.863 Cox& Snell R Square .077 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 164

164 Table D 4 Logistic regression model for victimization by psychological abuse in Korea Variables N=1399 Missing:124 B Valid N:1275 P OR Independent Variables Victim of c hild a buse .112** .001 1.118 Witnessed fathers violence .506 .745 1.057 Witnessed mothers violence .035 .888 .966 Self c ontrol scale .033** .000 .967 Attachment to mother .019 .419 .981 Attachment to father .014 .477 1.014 Control Variables Sex .503** .000 1.653 Age .179** .004 1.196 Family structure .599 Single parents .124 .703 1.132 Step parents .747 .206 2.112 Others .209 714 .812 Dating involvement .000 Occasionally 1.059** .000 2.884 Steady 1.202** .000 3.327 O thers .135 .757 .874 Campus .180 .570 .836 Risky s exual behavior .101** .000 1.106 Constant .239 .734 .787 Model Chi Sq 197.902 2LL 1416.764 Cox&Snell R Square .144 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 165

165 Table D 5 Logistic regression model for perpetration of physical violence in the U.S. Variables Valid N:1158 B P OR Independent Variables Victim of c hild a buse .126 ** .004 1.134 Witnessed fathers violence .145 .653 1.156 Witnessed mothers violence .004 .994 1.004 Self c ontrol scale .028 ** .001 .972 Attachment to mother .007 .759 1.007 Attachment to father .020 .258 .980 Control Variables Sex .875 ** .000 2.399 Age .056 .458 .945 Family structure .430 Single parents .059 .794 1.060 Step parents .421 .102 1.524 Others .075 .878 .927 Dating involvement .000 Occasionally .654 ** .004 1.924 Steady 1.033 ** .000 2.810 O thers 1.435 ** .000 4.201 Campus .083 .683 .920 Risky Sexual behavior .087 ** .000 1.091 Race .790 Hispanic .247 .268 .781 Black .089 .701 1.093 Asian .067 .844 1.069 Other .004 .994 1.004 Constant .688 .353 .503 Model Chi Sq 152.569 2LL 1146.493 Cox&Snell R Square .123 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 166

166 Table D 6 Logistic regression model for perpetration of psychological abuse in the U.S. Variables Valid N:1111 B P OR Independent Variables Victim of c hild a buse .130 ** .003 1.139 Witnessed fathers violence .058 .860 .943 Witnessed mothers violence .344 .261 1.411 Self c ontrol scale .017 .030 .983 Attachment to mother .012 .641 1.012 Attachment to father .009 .608 .991 Control Variables Sex .684 ** .000 1.982 Age .048 .505 .953 Famil y structure .253 Single parents .395 .072 .674 Step parents .138 .630 1.148 Others .188 .721 1.207 Dating involvement .000 Occasionally .996 ** .000 2.708 Steady 1.339 ** .000 3.816 O thers 1.842 ** .000 6.308 Campus .271 .096 1.312 Risky s exual behavior .160 ** .000 1.173 Race .314 Hispanic .398 .055 .672 Black .123 .589 .884 Asian .118 .716 .888 Other .528 .231 .590 Constant .756 .269 .470 Model Chi Sq 282.634 2LL 1252.999 Cox&Snell R Square .225 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 167

167 Table D 7 Logistic regression model for victimization by physical violence in the U.S. Variables Valid N:1161 B P OR Independent Variables Victim of c hild a buse .173 ** .000 1.189 Witnessed fathers violence .211 .535 1.235 Witnessed mothers violence .013 .966 1.013 Self c ontrol scale .029 ** .002 .971 Attachment to mother .009 .718 .991 Attachment to father .000 .986 1.000 Control Variables Sex .207 .244 1.230 Age .027 .742 1.027 Family structure .737 Single parents .224 .353 1.251 Step parents .063 .830 .939 Others .251 .654 .778 Dating involvement .000 Occasionally .665 ** .009 1.945 Steady 1.031 ** .000 2.805 Others 1.796 ** .000 6.026 Campus .104 .601 1.109 Risky s exual behavior .100 ** .000 1.106 Race .976 Hispanic .104 .663 .901 Black .087 .740 .917 Asian .109 .762 1.115 Other .122 .820 1.189 Constant 1.095 .168 .335 Model Chi Sq 134 100 2LL 987.361 Cox&Snell R Square .109 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 168

168 Table D 8 Logistic regression model for victimization by psychological abuse in the U.S. Variables Valid N:1118 B P OR Independent Variables Victim of c hild a buse .198 ** .000 1.220 Witnessed fathers violence .007 .983 .993 Witnessed mothers violence .023 .937 1.023 Self c ontrol scale .022 ** .005 .979 Attachment to mother .044 .065 1.045 Attachment to father .017 .312 1.017 Control Variables Sex .460 ** .002 1.584 Age .028 .695 .973 Family structure .621 Single parents .186 .386 .830 Step parents .163 .541 1.177 Others .318 .538 1.374 Dating involvement .000 Occasionally .980 ** .000 2.664 Steady .986 ** .000 2.679 Others 1.003 ** .004 2.726 Campus .161 .321 1.174 Risky s exual behavior .166 ** .000 1.181 Race .234 Hispanic .428 .037 .652 Black .328 .141 .721 Asian .041 .902 .960 Other .171 .706 .843 Constant 1.580 .018 .206 Model Chi Sq 246.362 2LL 1295.261 Cox&Snell R Square .198 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 169

169 Table D 9 Logistic regression result of full model -perpetration of physical violence Variables Valid N: 2456 B P OR Independent Variables Victim of c hild a buse .131** .000 1.140 Witnessed fathers violence .220 .196 1.247 Witnessed mothers violence .084 .686 .920 Self c ontrol scale .029** .000 .971 Attachment to mother .002 .918 .998 Attachment to father .022 .113 .979 Control Variables Country .460** .001 .1584 Sex .634** .000 1.885 Age .029 .586 .972 Family structure Single parents .073 .700 1.076 Step parents .433 .066 1.541 Others .151 .699 1.163 Dating involvement Occasionally .933** .000 2.543 Steady 1.003** .000 2.728 Others 1.267** .000 3.551 Campus .182 .228 .833 Risky s exual behavior .091** .000 1.095 Constant .842 .107 .431 Model Chi Sq 293.046 2LL 2133.557 Cox&Snell R Square .112 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 170

170 Table D 10 Logistic regression result of full model perpetration of psychological abuse Variables Valid N: 2383 B P OR Independent Variables Victim of c hild a buse .127** .000 1.136 Witnessed fathers violence .177 .236 1.194 Witnessed mothers violence .020 .914 .980 Self c ontrol scale .018** .002 .982 Attachment to mother .022 .179 .982 Attachment to father .000 .996 1.000 Control Variables country .882** .000 2.417 Sex .568** .000 1.765 Age .071 .129 1.073 Family structure Single parents .205 .243 .815 Step parents .147 .556 1.158 Others .115 .758 .891 Dating involvement Occasionally .929** .000 2.532 Steady 1.172** .000 3.230 Others 1.024** .000 2.784 Campus .095 .481 1.099 Risky s exual behavior .143** .000 2.417 Constant 1.113* .012 .329 Model Chi Sq 551.810 2LL 2682.563 Cox&Snell R Square .207 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 171

171 Table D 11 Logistic regression result of full model victimization by physical violence Variables Valid N: 2463 B P OR Independent Variables Victim of c hild a buse .150** .000 1.162 Witnessed fathers violence .146 .427 1.157 Witnessed mothers violence .075 .727 1.078 Self c ontrol scale .036** .000 .965 Attachment to mother .033 .086 .968 Attachment to father .006 .691 .994 Control Variables Country .521** .001 1.684 Sex .381** .004 1.464 Age .021 .716 1.021 Family structure Single parents .138 .498 1.148 Step parents .058 .824 1.060 Others .011 .979 .989 Dating involvement Occasionally .740** .000 2.096 Steady .913** .000 2.491 Others 1.347** .000 3.844 Campus .065 .703 1.067 Risky s exual behavior .097** .000 1.102 Constant .615 .278 .541 Model Chi Sq 240.821 2LL 1864.031 Cox&Snell R Square .093 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 172

172 Table D 12 Logistic regression result of full model victimization by psychological abuse Variables Valid N: 2386 B P OR Independent Variables Victim of c hild a buse .141** .000 1.151 Witnessed fathers violence .038 .799 1.039 Witnessed mothers violence .000 1.000 1.000 Self c ontrol scale .026** .000 .975 Attachment to mother .007 .656 1.014 Attachment to father .014 .250 .975 Control Variables Country .397** .001 1.487 Sex .491** .000 1.634 Age .088 .053 1.092 Family structure Single parents .100 .561 .904 Step parents .224 .345 1.251 Others .095 .796 .1.100 Dating involvement Occasionally 1.023** .000 2.782 Steady 1.101** .000 3.008 Others .547* .032 1.728 Campus .046 .732 1.047 Risky s exual behavior .134** .000 1.143 Constant 1.156** .009 .315 Model Chi Sq 454.180 2LL 2734.802 Cox&Snell R Square .173 *p<.05. **p<.01. reference group: two biological parents reference group: currently not dating

PAGE 173

173 LIST OF REFERENCES Alexander, P. C., Moore, S., & Alexander, E. (1991). What is transmitted in the intergenerational transmission of violence? Journal of Marriage and Family, 53, 657668. Arias, I., Samios, L., & OLeary, K. (1987). Prevalence and correlates of physical aggression during courtship. Journal of Interpersonal Violence 2 82 90. Arias, I, & Johnson, P. (1989). Evaluations of physical aggression among intimate dyads. Journal of Interpersonal Violence, 4 298 307 Avakame, E. F. (1998a). Intergenerational transmission of violence, self -control, and conjugal violence: A comparative analysis of physical violence and psychological aggression. Violence and Victims, 13, 301316. Avakame, E. F. (1998b). Intergenerational transmission of violence and psychological aggression against wives. Canadian Journal of Behavioral Science 30, 193202. Bandura, A. (1965). Influence of models reinforcement contingencies in the acquisition of imitative responses. Journal of Personality and Social Psychology, 1 589 595. Bandura, A. (1973). Aggression: A social learning analysis Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1977). Self -efficacy: Toward a unified theory of behavioral change. Psychological Reviews 84, 191215. Bergman, L. (1992). Dating violence among high school students. Social Work, 37, 21 27. Bernard, M. & Bernard, J. (1983). Violent intimacy: The family as a model for love relationships. Family Relations, 32, 283 286. Bethke, T., & DeJoy, D. (1993 ). An experimental study of factors influencing the acceptability of dating violence. Journal of Interpersonal Violence, 8 36 51. Billingham, R. (1987). Courtship violence: The patterns of conflict resolution strategies across seven levels of emotional co mmitment. Family Relations, 36, 283 289. Billingham, R., & Gilbert, K. (1990). Parental divorce during childhood and use of violence in dating relationships. Psychological Reports 66 1003 1009. Billingham, R., & Notebaert, N. (1993). Divorce and dating violence revisited: Multivariate analyses using Straus Conflict Tactics subscores. Psychological Reports 73 679 684. Banyard, V. L., Cross, C. &Modecki, K. L (2006. Ecological Correlates of Self Reported Perpetration Journal of Interpersonal Violence 21, 13141332. Bergman, L. (1992). Dating violence among high school students. Social Work, 37, 21 27

PAGE 174

174 Bernard, J. L., Bernard, S. L., & Bernard, M. L. (1985). Courtship violence and sex typing Family Relations 34, 573576. Brener, N.D., McMahon, P. M., Warren, C. W., & Douglas, K. A. (1999). Forced sexual intercourse and associated health risk behaviors among female college students in the United States. Journal of Consulting and Clinical Psychology, 67, 252259. Bowlby, J. (1973). Attachment and los s: Vol.2, Separation. New York: Basic Cassidy, J. (1994). Emotion regulation: Influences of attachment relationships. Monographs of the Society for Research in Child Development 59, 228283. Cassidy, J. (2000). Adult romantic attachments: A developmental perspective on individual differences. Review of General Psychology 4 11 131. Cate, R. M., Henton, J.M., Koval, J., Christopher, F. S., & Lloyd, S. (1982). Premarital abuse: A social psychological perspective. Journal of Family Issues 3, 79 90 Carr, J., & VanDeusen, K. M. (2002). The relationship between family of origin violence and dating violence in college men Journal of Interpersonal Violence, 17, 630646. Cicchetti, D., & Howes, P. W. (1991). Developmental psychopathology in the context of the family: Illustrations from the study of child maltreatment. Canadian Journal of Behavioral Science 23, 257 281. Centers for Disease Control. Dating Abuse Fact Sheet http://www.cdc.gov/ncipc/dvp/DatingViolence.htm Accessed on August 10, 2006. Chapple, C. L., & Hope, T. L. (2003). An analysis of the self -control and criminal versatility of gang and dating violence offenders. Violence and Victims, 18, 671690. Chase, K. A., Treboux, D., & OLeary, K. D. (2002). Characteristics of high risk adolescents dating violence. Journal of Interpersonal Violence, 17, 3349. Cleveland, H., Herrera, V. M. & Stuewig, J. (2003). Abusive males and abused females in adolescent relationships: Risk factor s imilarity and dissimilarity and the role of relationship seriousness. Journal of Family Violence 18, 325339. Crittenden, P.M & Ainsworth, D.S. (1989). Child maltreatment theory Cicchetti & Carlson (eds), Handbook of child maltreatment (pp. 432246). Cam bridge: Cambridge University Press. Dodge, K. A., Petit, G. S., Bates, J. E., & Valente, E. (1995). Social information processing patterns mediate the effect of early physical abuse on later conduct problems. Journal of Abnormal Psychology 104, 632 643. Eisikovits, Z., Winstok, Z., & Fisher, G. (2004). The first Israeli national survey on domestic violence. Violence Against Women, 10, 729748.

PAGE 175

175 Coffey, P., Leitenberg, H., Henning, K., Bennett, R., & Jankowski, M. (1996). Dating violence: The association between methods of c oping and womens psychological adjustment Violence and Victims, 11, 227 238. Ferguson, D.M. ,Boden, J.M. & Horwood, L.J. Circumcision Status and Risk of Sexually Transmitted Infection in Young Adult Males: An Analysis of a Longitudinal Birth Cohort. Pediatrics (2006), 118, 19711977. Follette, V. (1992). The effect of violence in the family of origin on problem -solving strategies in dating couples. Behavioral Assessment, 14, 1 3. Follette, V., & Alexander, P. (1992). Dating violence: Current and historical correlates. Behavioral Assessment, 14, 39 52. Follingstad, D., Rutledge, L., Polek, D., & McNeill Hawkins, K (1988). Factors associated with patterns of dating violence toward college women. Journal of Family Violence, 3 169 182. Fo llingstad, D. R., Bradley, R. G., Laughlin, J. E., & Burke, L. (1999). Risk factors and correlates of dating violence: The relevance of examining the frequency and severity levels in a college sample. Violence and Victims, 14, 365 380. Higgins, G. E. (2007). Examining the original Grasmick scale: A Rasch model approach. Criminal Justice and Behavior, 34,157178. Holtfreter, K. Reisig, M., & Pratt T. (2008). Low self -control, routine activities, and fraud victimization. Criminology 46 (1): Accepted & forthcoming. Foo, L, & Margolin, G. (1995). A multivariate investigation of dating aggression. Journal of Family Violence, 10 351 377. Foshee, V. (1996). Gender differences in adolescent dating abuse prevalence, types, and injuries. Health Education Research, 11, 275 286. Foshee, V. Bauman, K E., & Linder, G. F. (1999). Family violence and the perpetration of adolescent dating violence: Examining social learning and social control processes. Journal of Marriage and the Family, 61, 331342. Foshee, V., Be nefield, T., Ennett, S., Bauman, K., & Suchindran, C.(2004). Longitudinal predictors of serious physical and sexual dating violence victimization during adolescence. Preventive Medicine 39, 1007 1016. Foshee, V., Bauman, K., Linder, F., Rice, J., & Wilche r, R. (2007).Typologies of adolescent dating violence: Identifying typologies of adolescent dating violence perpetration. Journal of Interpersonal Violence 22, 489 519. Goldstein, D., & Rosenbaum, A. (1985). An evaluation of the self -esteem of maritally v iolent men. FamilyRelations, 34, 425 428.

PAGE 176

176 Gottfredson, M. & Hirschi, T. (1990). A general theory of crime Palo Alto: CA. Stanford University Press. Gover, A. R. (2004). Risky lifestyles and dating violence: A theoretical test of violent victimization. Jou rnal of Criminal Justice, 32, 171180. Gover, A. R. Kaukinen, C. & Fox, K. (2008) The Relationship Between Violence in the Family of Origin and Dating Violence Among College Students Journal of Interpersonal V iolence 23, 16671693. Grasmick, H. G., Tittle,C., Bursik, R. J., & Arneklev, B. K.(1993) Testing the Core Empirical Implications of Gottfredson and Hirschis General Theory of Crime. Journal of Research in Crime and Delinquency 30 5 29. Gray, H., & Foshee, V. (1997). Adolescent dating violence: Differences between one -sided versus mutually violent profiles. Journal of Interpersonal Violence 12 126 141. Gross, A. M., Winslett, A., Roberts, M. ,& Gohm, C. L (2006) An examination of sexual violence against college women. Violence Against Women 12, 288300. Gwartney Gibbs, P., Stockard, J., & Brohmer, S. (1987). Learning courtship violence: The influence of parents, peers, and personal experiences. Family Relations Journal of Applied Family and Child Studies 36 276 282. Hanley, M. J., & ONei ll, P. (1997). Violence and commitment: A study of dating couples. Journal of Interpersonal Violence, 12, 685 703. Harned, M. S.(2002). A multivariate analysis of risk marker for dating violence victimization. Journal of Interpersonal Violence 17 11791197. Henton, J. M., Cate, R. M., Koval, J., Lloyd, S., & Christopher, F. S. (1983). Romance and violence in dating relationships Journal of Family Issues 4, 467482. Hickman, L. J Jaycox L. H., & Aronoff, J (2004) Dating Violence among Adolesc ents. Trauma, Violence, & Abuse 5 123142. Hirschi T (2004) Self-control and crime In: Baumeister RF, Vohs KD (eds) Handbook of self regulation. Guilford Press, New York Holtzworth -Munroe, A., & Stuart, G. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin 116, 476 497 Jackson, S. M. (1999). Issues in the dating violence research: A review of the literature. Aggression and Violent Behavior 4, 233 247. Jaffe, P.G., Wolfe, D.A., & Wilson, S.K. (19 90). Children of Battered Women, Sage. Newbury Park, CA.

PAGE 177

177 Jankowski, M. K., Leitenberg, H., Henning, K., & Coffey, P. (1999). Intergenerational transmission of dating aggression as a function of witnessing only same sex parents vs. opposite sex parents vs. both parents as perpetrators of domestic violence. Journal of Family Violence, 14, 267279. Kaufman, J. and Zigler, E. (1993), The Intergenerational Transmission of Abuse is Overstated, in R. J. Gelles and D. R. Loseke (eds), Current Controversies on Family Violence Newbury Park, Calif.; London, Sage. Kerley, K. R., Xu, X., & SirisunyaluckB. (2008) Self Control, Intimate Partner Aggression, and Intimate Partner Victimization: Testing The General Theory of Crime in Thailand Deviant Behavio r, 29, 503 532. Kwong, M. J., Bartholomew, K., Henderson, A. J. Z., & Trin ke, S. J. (2003). The intergenerational transmission of relationship violence. Journal of Family Psychology, 17, 288301. Lewis, S. F., & Fremouw, W. (2000). Dating violence: A critical review of the literature. Clinical Psychology Review 21, 105127. Luthra, R. &. Gidycz G A (2006). Dating Violence Among College Men and Women. Journal of Interpersonal Violence 21,717731 Lysova, A. V. (2007). Dating violence in Russia. Russian Education and S ociety, 49, 43 59. Lloyd, S. (1991). The dark side of courtship: Violence and sexual exploitation. Family Relations, 40, 14 20. Magdol, L., Moffitt, T., Caspi, A., Newman, D., Fagan, J., & Silva, P. (1997). Gender differences in partner violence in a birth cohort of 21-year olds: Bridging the gap between clinical and epidemiological approaches. Journal of Consulting and Clinical Psychology, 65, 68 78. Magdol, L., Moffitt, T., Caspi, A., & Silva, P. (1998). Hitting without a license: Testing explanations for differences in partner abuse between young adult daters and cohabitators. Journal of Marriage and the Family, 60, 41 55. Makepeace, J. M. (1981). Courtship violence among college students. Family Relations, 30, 97 102. Makepeace, J. M. (1983). Life events, stress, and courtship violence. Family Relations, 32, 101 109. Makepeace, J. M. (1986). Gender differences in courtship violence victimization. Family Relations, 35, 383 388 Makepeace, J. M. (1987). Social factors and victim offender differences in courtship violence. Family Relations, 36, 87 91.

PAGE 178

178 Marshall, L, & Rose, P. (1987). Gender, stress, and violence in adult relationships of a sample of college students. Journal of Social and Personal Relationships 4 299 316. Marshall, L. L., & Rose, P. (1988). Family of origin violence and courtship abuse. Journal of Counseling and Development, 66, 414418. Mawell, C. D. & Maxwell, S. R. (2003). Experiencing and witnessing familial aggression and their relationship to physica lly aggressive behaviors among Filipino adolescents. Journal of Interpersonal Violence, 18, 14321451. Migcot, M.,& Lester, D. (1996). Psychological abuse in dating, locus of control, depression, and suicidal preoccupation. Psychological Rep 79, 682. Mills. C.S., & Granoff, B.J. (1992). Date and acquaintance rape among a sample of college students. Social Work 27, 504509. Monson, C. M., & LanghinrichsenRohling, J. (2002). Sexual and nonsexual dating violence perpetration. Violence and Victims, 17, 403428. Muehlenhard, C. L., & Linton, M. A. (1987). Date rape and sexual aggression in dating situations: Incidence and risk factors. Journal of Counseling Psychology, 34, 186196. OKeefe, M. (1997). Predictors of dating violence among high school stude nts. Journal of Interpersonal Violence 12, 546568. OKeefe, M. (1998). Factors mediating the link between witnessing interparental violence and dating violence. Journal of Family Violence 13, 3957. OKeefe, M. & Treister, L. (1998). Victims of dating violence among high school students: Are the predictors different for males and females? Violence Against Women 4, 195223. OLeary, K. D. (1999). Psychological abuse: A variable deserving critical attention in domestic violence. Violence and Victims 14, 3 23. Paternoster, B., Brame,R., Mazerolle, P. & Piquero, A.(1998). Using the correct statistical test for the equality of regression coefficient Criminology 36, 859 866 Raj, A. & Silverman, J. G. (2003). Immigrant South Asian women at greater risk for injury from intimate partner violence. American Journal of Public Health, 93, 435437. Ramisetty -Mikler, S., Goebert, D., Nishimura, S., & Caetano, R. (2006). Dating violence victimization: Associated drinking and sexual risk behaviors of Asian, Native Hawaiian, and Caucasian high school students in Hawaii. Journal of School Health, 76, 423 429. Reuterman, N. A., & Burcky, W.D. (1989). Dating violence in high school: A profile of the victim. Psychology Journal of Human B ehavior 26, 19.

PAGE 179

179 Riggs, D., OLeary, K. & Breslin, F. (1990). Multiple correlates of physical aggression in dating couple s. Journal of Interpersonal Violence, 5, 61 73. Riggs, D. (1993). Relationship problems and dating aggression. Journal of Interpersonal Violence, 8 18 35. Riggs, D., & OLeary, K. (1996). Aggression between heterosexual dating partners. Journal of Interp ersonal Violence 11 519 540. Riggs, D., & Caulfield, M. (1997). Expected consequences of male violence against their female dating partners. Journal of Interpersonal Violence, 12, 229 240. Riggs, D. S., Caulfield, M. B., & Street, A. E. (2000). Risk for domestic violence: Factors associated with perpetration and victimization. Journal of Clinical Psychology, 56, 12891316. Rowe, D.C., Vazsonyi, A. T., & Figueredo, A.J. (1997). Mating effort in adolescence: A conditional or alternative strategy. Personal Individual Differences 23, 105115. Schiff, M. & Zeira, A. (2005). Dating violence and sexual risk behaviors in a sample of at risk Israeli youth. Child Abuse & Neglect, 29, 1249 1263. Schreck, C., Stewart E. A. & Fisher, B.S. (2006).Self -control, vic timization, and their influence on risky lifestyles: A longitudinal analysis using panel data. Journal of Quantitative Criminology 22,319340. Schwartz J. P., Hage, S. M., Bush, I., & Bruns, L K. (2006). Unhealthy Parenting and Potential Mediators as Contributing Factors to Future Intimate Violence: A Review of the Literature T rauma, Violence, & Abuse 7 206221 Sellers, C. (1999). Self -control and intimate violence: An examination of the scope and specification of the general theory of crime. Criminolo gy, 37, 375404. Sellers C., Cochran J. & Branch K. ( 2005). Social Learning Theory and Partner Violence: A Research Note. Deviant Behavior 26, 379395 Silverman, J. G., Raj, A., Mucci, L. A., & Hathaway, J. E. (2001). Dating violence against adolescent girls and associated substance use, unhealthy weight control, sexual risk behavior, pregnancy, and sociality. Journal of the American Medical Association, 286, 572579. Sharpe, D., & Taylor, J. K( 1999). An examination of variables from a social -developmental model to explain physical and psychological dating violence. Canadian Journal of behavioral Science 31,165175 Smith, J., & Williams, J. (1992). From abusive household to dating violence. Jour nal of Family Violence, 7 153 165.

PAGE 180

180 Smith P H, White J W & Holland, L J .( 2003). A Longitudinal Perspective on Dating Violence Among Adolescent and College -Age Women. American Journal of Public Health; 93 11041109. Sternberg, K. J., Lamb, M. E., Guterman, E., Abbott. C.B., & Dawud Noursi, S. (2005) Adolescents perceptions of attachment to their mothers and fathers in families with histories of domestic violence: a longitudinal perspective Child Abuse Neglect 29, 853869. Sroufe, L. A., Carlson, E. A., Levy, A. K., & Egeland, B. (1999). Implications of attachment theory for developmental psychopathology. Development and Psychopathology 11, 1 13. Sternberg, K. J., Lamb, M. E., Greenbaum, C., Cicchetti, D., Dawud, S., Cortes, R. M., & Lorey, F. (1 993). Effects of domestic violence on childrens behavior problems and depression. Developmental Psychology 29, 44 52. Sternberg, K. J., Lamb, M. E., Greenbaum, C., Dawud, S., Cortes, M., & Lorey, F. (1994). The effects of domestic violence on childrens perceptions of their perpetrating and nonperpetrating parents. International Journal of Behavioral Development 17, 779 795. Stets, J., & Straus, M. (1989). The marriage license as a hitting license: A comparison of assaults in dating, cohabiting, and married couples. Journal of Family Violence 4 161 180. Stith, S., Jester, S., & Bird, G. (1992). A typology of college students who use violence in their relationships. Journal of College Stude nt Development, 33 411 421. Straus, M. A., & Gelles, R. J. (1986). Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and the Family, 41, 75 88. Straus, M. A., Hamby, S. L., Boney McCoy S., & Sugarman, D. B. (1996). The revised conflict tactics scales (CTS2): Development and preliminary psychometric data. Journal of Family 17, 283 316. Straus, M. A. (2004). Prevalence of violence against dating partners by male and female university stu dents worldwide. Violence Against Women, 10, 790811. Sugarman, D., & Hotaling, G. (1989). Dating violence: Prevalence, context, and risk markers. In M. Pirog Good & J. Stets (Eds.), Violence and dating relationships (pp. 3 32). New York: Praeger. Sugarman, D.A., & Hoteling,G. T. (1989). Dating violence: Prevalence, context, and risk makers. In Pirog Good, M.A., & Stets, J.E. (eds), Violence in Dating Relationships: Emerging Social Issues Praeger, New York, Pp. 3 3 Tontodonato, P., & Crew, B. (1992). Dating violence, social learning theory, and gender: A multivariateanalysis. Violence and Victims 7 3 14.

PAGE 181

181 Watson, J. M., Cascardi, M., Avery Leaf, S., & OLeary, K. D. (2001). High school students responses to dating aggression. Violence and Victims, 16, 339 348. White, J. W., & Koss, M. P. (1991). Courtship violence: Incidence in a national sample of higher education students. Violence and Victims 6, 247256.

PAGE 182

182 BIOGRAPHICAL SKETCH MiRang Park grew up in Seoul, Korea. After graduation from high school she obtained her bachelors degree in Journalism with minor in sociology from Ewha Womans University in Seoul, Korea. After completing her undergraduate degree in 2003, she moved to East Lansing, Michigan and later her masters d egree in criminal justice from the Michigan State University. In 2009 she completed her Ph.D in criminology, law and society at the University of Florida in Gainesville, Florida. Her interests include 10 mile running, kayaking, traveling, and rice noodle She is still young and beautiful.

PAGE 183

183 FAMILY EXPERIENCE, ATTACHMENT TO PARENTS AND LOW SELF CONTROL IN DATING VIOLENCE IN THE U.S. AND KOREA MiRang Park 3522620949 Department of Sociology, Criminology and Law & Society Supervisory chair : Ronald Akers Co -char: Angela Gover Ph.D May, 2009 Although many studies have been done about dating violence, only a few researchers have developed theoretical conceptualizations of factors such as these in dating violence (Sellers, 2005; Follingstad et a l, 1999; Riggs and OLeary, 1989; Luthra and Gidycz, 2006). This dissertation moves beyond these to draw on three prominent general theories of crime and deviance for its theoretical framework to understand the effects of early family experience with viole nce (social learning), current attachment to parents (social bonding), and self -control. The purpose of this study is to link theoretically relevant variables that are linked in some way to the respondents family of origin that may account for the likel ihood of perpetration of or victimization by dating violence in two different cultures. In using the same survey instrument for two countries, this study tested whether theoretical explanations developed in Western society could be applied to Eastern soci ety. A sample of 1,400 young adults attending a university in Korea and a sample of 1,500 young adults attending a university in America were used in this dissertation. Findings indicate that experience of victim of child abuse (social learning), low self -control (self -control), serious dating relationship and risky sexual behavior have significant effects on both the perpetration and victimization of dating violence in both countries. Self -control has stronger effects in the American, and being the victim of violence in the home has more significant effects in the Korean sample.