INFORMATION TO USERS This manuscript has been reproduced from the microfilm m aster. UMI films the text directly from the original or copy submitted. Thus, som e thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality o f th is reproduction is dependent upon th e quality o f the copy subm itted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand com er and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. ProQuest Information and Learning 300 North Zeeb Road, Ann Aitor, Ml 48106-1346 USA 800-521-0600 UMI SEX-ROLE BELIEFS OF CHILDREN EXPOSED TO WIFE ABUSE by Kendra Rogers B.A., The University o f B ritish Columbia, 1993 Diploma in Guidance Studies, The University o f British Columbia, 1995 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF EDUCATION in COUNSELLING © Kendra Rogers, 1998 THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA April 1998 All rights reserved. This work may not be reproduced in whole or in part, by photocopy or other means, without the perm ission o f the author. 1^1 National Library of Canada Bibliothèque nationale du Canada Acquisitions and Bibliographic Services Acquisitions et senrices bibliographiques 395 Wellington Street Ottawa ON K1A0N4 Canada 395, rue WeVngion Ottawa ON K1A0N4 Canada Yourm Yo09rmtWK9 Our§Ê9 Notnréténnc» The author has granted a non­ exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of this thesis in microform, paper or electronic formats. L’auteur a accordé une licence non exclusive permettant à la Bibliodièque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfîche/fîlm, de reproduction sur papier ou sur format électronique. The author retains ownership of the copyright in this thesis. Nei&er the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author’s permission. L’auteur conserve la propriété du droit d’auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation. 0-612-62494-3 CanadS APPROVAL Name: Kendra Rogers Degree: Master o f Education in Counselling Thesis Title: SEX-ROLE BELIEFS OF CHILDREN EXPOSED TO WIFE ABUSE Examining Committee: Chair: Dr. Martha MacLeod Associate Professor and Program Chair, Nursing Program, UNBC; Associate Professor and Program Co-Chair, Community Health Program,UNBC Supervisor: Dr. Peter MacMillan Assi^ant Professo^ Education Program, UNBC Dr. Ron Lehr Assistant Professor, Education Program, UNBC Ms. Barbara Isaac Assistant Professor, SqciaLWork Program, UNBC jordon Temowetsky Professor, Social Work Program, UNBC Date Approved: u ABSTRACT Extensive research exists on the effects o f wife abuse on its fem ale victims, but only recently has attention been directed to the children exposed to w ife abuse in their homes. Children exposed to wife abuse display a wide range o f physical, emotional, behavioural, and cognitive problem s. However, there are also numerous invisible effects o f exposure to wife abuse which have not been extensively researched. One such effect m ay be the children’s incorporation o f the traditional sex-role beliefs often present in fam ilies characterized by wife abuse. Research shows that m en who abuse their wives often hold traditional beliefs about their right to control and dominate their partner. After prolonged abuse at the hands o f a dom inant and controlling man, women may view themselves as powerless and weak. If children exposed to wife abuse incorporate these differential beliefs about power and control, they may be more likely to become involved in abusive relationships as adults. Specifically, boys may be more likely to hold attitudes condoning wife abuse, and girls may believe they can not prevent wife abuse firom starting, or stop their victim ization if w ife abuse does occur. This study included a treatm ent group o f 12 children exposed to wife abuse and a comparison group o f 12 children not exposed to wife abuse. The children completed the Child Sex Role Inventory (CSRl), a questionnaire based on the Bem Sex Role Inventory (Bern, 1981). The CSRl assesses the degree to which the children describe themselves as traditionally m asculine or feminine sex-typed (Boldizar, 1991). The children’s mothers com pleted the Bem Sex Role Inventory (1981), the Conflict Tactics Scales (CTS) (Straus, 1979), an adapted version o f the CTS which assessed their children’s exposure to wife abuse. Ill and a demographic questionnaire. The children exposed to wi& abuse w ere compared to the children not exposed to wife abuse on their exposure to wife abuse and their self-reported sex-role beliefs. The results indicated that children exposed to wife abuse rated themselves as somewhat more traditionally sex-typed than children not exposed to wife abuse. Boys in the treatm ent and comparison groups did not differ significantly in the extent to which they described themselves as masculine, but boys exposed to wife abuse described themselves as significantly less feminine than boys not exposed to wife abuse. Girls exposed to wife abuse did not differ significantly fix)m girls not exposed to wife abuse on either their self-reported m asculinity or femininity. The implications o f the findings for fam ily violence researchers, counsellors, parents, teachers, and social workers are discussed, and recommendations are made for fiiture research. IV TABLE OF CONTENTS Abstract ü Table o f Contents iv List o f Tables vii L ist o f Figures vüi Acknowledgment ix CHAPTER ONE: INTRODUCTION 1 Definition o f Terms 3 Scope and Lim itations o f Study 6 Outline o f Thesis 6 CHAPTER TW O: LITERATURE REVIEW 8 W ife Abuse 8 Theoretical Analyses o f W ife Abuse 10 Psychological Theory 10 Sociological Theory 11 Socialization o f Gender Theories 13 Fem inist Theory 15 Abusive Personality Theory 18 Theoretical Perspective o f Current Study 19 Sex-Role Beliefs o f M ale Abusers and Female Victims 20 Estimates o f Children Exposed to W ife Abuse 22 Effects o f Children’s Exposure to W ife Abuse 23 Sex-Role Beliefs and the Intergenerational Transmission o f Violence Theory 26 Rationale for Present Study 31 Hypotheses 32 CHAPTER THREE: METHOD 34 Recruitm ent o f Participants 34 Independent Variable 37 Instrum ents 39 Demographic Questionnaire 40 Conflict Tactics Scales 40 Adapted Version o f the Conflict Tactics Scales 43 Sex Role Inventories 43 Setting 47 Procedure 48 Collection, Recording, and Analysis o f D ata 49 M ethodological Assumptions 50 Lim itations SO CHAPTER FOUR: RESULTS 53 Participants 53 Estim ates o f Children’s Exposure to W ife Abuse 57 BSRI and CSRl Classification o f M asculinity and Femininity 59 D ata A nalysis 60 W omen’s BSRI Scores 60 Children’s CSRl Scores 63 VI Summary o f Results 68 CHAPTER FIVE: CONCLUSIONS 71 Summary o f Literature Review 71 Summary o f M ethods 71 Summary o f Results 72 Conclusions 72 W omen’s BSRI Scores 72 Children’s CSRl Scores 73 Explanation o f F hidings 74 W omen’s BSRI Results 74 Children’s CSRl Results 75 Integration o f Findings with Past Literature 78 Im plications o f Findings 79 Lim itations 82 Future Directions 84 Footnotes 87 References 88 Appendix A: Instruments 93 Appendix B: Letters for Recruitment o f Treatment Group 103 Appendix C: Letters for Recruitment o f Comparison Group 108 Appendix D: Instructions for Children’s Completion o f the CSRl 112 A ppendix E: Consent Forms 113 vil L ist o f Tables Table 1: Behavioural, Physical, and Psychological Effects o f Exposure to W ife Abuse 24 Table 2: Num ber o f Treatment Group W omen and Children from Each Agency 36 Table 3: Sex-Role Dimensions o f the BSRI 44 Table 4: Sample Items from the BSRI and CSRl 47 Table 5: Characteristics o f M others o f Treatment & Comparison Group Children 56 Table 6: Treatm ent and Comparison Groups’ Exposure to Wife Abuse 58 Table 7: Sex-Role Categories o f the CSRl 59 Table 8: BSRI M asculinity Scores o f Treatment vs. Comparison Group M others 61 Table 9: BSRI Femininity Scores o f Treatm ent vs. Comparison Group M others 62 Table 10: Summary o f Treatment and Comparison Group M others’ BSRI Results 62 Table 11: CSRl M asculinity Scores o f Boys EWA vs. Boys NEWA 64 Table 12: CSRl Femininity Scores o f Boys EWA vs. Boys NEWA 64 Table 13: CSRl M asculinity Scores o f Girls EWA vs. Girls NEWA 66 Table 14: CSRl Femininity Scores o f Girls EWA vs. Girls NEWA 66 Table 15: CSRl M asculinity Scores o f Boys EWA vs. Girls EWA 68 Table 16: CSRl Femininity Scores o f Boys EWA vs. Girls EWA 68 Table 17: Summary o f Children’s CSRl Results 69 V lll List o f Figures Figure 1: Figure 2: Figure 3 : Figure 4; Graphic representation o f the apparent disordinal interaction between boys’ m asculinity and fem ininity scores on the C SR l. 65 Graphic representation o f the apparent ordinal interaction between girls’ m asculinity and fem ininity scores on the CSRl. 67 Graphic representation o f the apparent disordinal interaction between group and m asculinity scores. 77 Graphic representation o f the apparent ordinal interaction between group and fem ininity scores. 77 IX ACKNOWLEDGMENT I would like to thank the three community mental health agencies and the elementary school firom which I gathered m y participants, as they enabled me to develop and conduct my study in comfortable and supportive enviromnents. I sincerely appreciated the fam ilies who contributed their tim e and thought to this study. The women in the treatm ent group shared parts o f their lives w hich were private and painful, and I admired their willingness to reveal experiences and personal inform ation in order to contribute to our knowledge o f the effects o f exposure to wife abuse. I also thank the wom en and children in the com parison group, as they participated w ith little to gain but a greater understanding o f them selves and their children. Every graduate student relies on the advice, suggestions, and support o f her com m ittee. I thank Ron, Barbara, and Gordon for their thorough and insightful reading o f and reflection on my thesis. I give special thanks to Dr. Peter MacM illan, whose faith in my abilities as a student and junior statistician was constant motivation. His support throughout both m y graduate courses and thesis supervision is much appreciated. I have been fortunate throughout my life to have the unconditional support o f my parents, firiends, and extended family. Thank you to all o f you for your encouragement and generosity - 1 could not have reached this goal without you. My partner and best fiiend, M auro Calabrese, has been invaluable to me during the past three years. I thank him for accepting his position as second fiddle to my academic goals and for rem inding me that there is life beyond grad school. CHAPTER ONE: SEX-ROLE BELIEFS OF CHILDREN EXPOSED TO WIFE ABUSE Throughout the family violence literature, wife abuse is referred to as domestic violence, m arital violence, spouse abuse, battering, and violence against women. I have chosen to use the term wife abuse throughout this study to reflect the fact that between 91 and 95 percent o f incidents o f spousal assault involve men’s abuse o f women (McCue, 1995). Wife abuse is a social problem which occurs in every type o f male-female relationship, including common-law and marital relationships. W hile the prevalence o f recent reports o f wife abuse may suggest it is a new problem, wife abuse is by no means a contemporary phenomenon. Centuries ago, men were permitted to and even encouraged to use violence against their wives to m aintain power over them (Johnson, 1996; W alker, 1979). Early marriage laws gave m en the legal right to hit their wives (Dobash & Dobash, 1992; M ullender & M orley, 1994). The frequently used phrase “rule o f thumb” originated as a British law lim iting husbands’ rights to hitting their wives with a rod no thicker than their thumb (Sigler, 1989). Despite changes to laws, cultural beliefs, and social attitudes, m en’s domination and control o f women through wife abuse continues today. Estimates o f the number o f Canadian women abused annually vary considerably, but numbers between 200,000 and 450,000 are common (Copping, 1996; Johnson, 1996). These countless women are emotionally, physically, and sexually harmed by such abuse. I f the women have children, they also suffer the consequences. It has been estim ated that children are exposed to between 39 and 80 percent o f wife abuse incidents (Health and W elfare Canada, 1992; W olfe & Jaffe, 1991). These numbers add up to between two and three m illion Canadian children being exposed to wife abuse every year (Johnson, 1996; Jaffe, W olfe, & W ilson, 1990). W itnessing specific violent acts is only one aspect o f children’s exposure to wife abuse. W hile children are seeing and hearing the abuse o f their mothers, they are also being taught a powerful and fiightening lesson: people who love each other m ay also hurt each other (Jaffe, W olfe, & W ilson, 1990). Traditional beliefs o f m ale dominance are common in fam ilies characterized by wife abuse. Abusive men are often dom inant and aggressive, and believe they have the right to exercise power and control over their wives. Abused women are firequently perceived as subm issive and powerless, and may become so in the face o f continued abuse. These behaviours, when repeated by the children in violent fam ilies, have far-reaching consequences for the children’s relationships as adults. Long-term exposure to traditional sex-role beliefs affects children in numerous ways. M others’ and fathers’ m odelling o f traditional sex-role behaviours increases children’s tendency to display sim ilar patterns o f behaviour (Barnett, M iller-Perrin, & Perrin, 1997; Choice, Lamke, & Pittman, 1995; C elles & Cornell, 1990). Traditional sex-role beliefs affect all human relationships, even in childhood. Children who hold traditional sex-role beliefs may develop relationships characterized by unequal power. Their peer relationships m ay then be vulnerable to conflict and struggles to assert or m aintain power. Such relationships, when carried into adolescent or adult life, have the potential to become violent. Although they are not predeterm ined to be abusive, children who are exposed to wife abuse are more likely to becom e abusive as adults ^ a m e tt et al., 1997; Choice et al., 1995; Celles & Cornell, 1990). This study is based on the hypothesis that traditional sex-role beliefs w ill be more prevalent in children exposed to wife abuse than in children who have not been exposed to such abuse. Samples are used, but it is the population that is o f interest. Specifically, it is hypothesized that m asculinity and femininity scores on the Child Sex Role Inventory (CSRl) (See Appendix A) w ill show that boys exposed to wife abuse view themselves as dominant, aggressive, and forcefiil, while girls exposed to wife abuse view themselves as passive, yielding, and compromising. An investigation o f these hypotheses w ill enable counsellors, parents, and the community to better understand and assist children exposed to wife abuse in their attem pts to avoid involvem ent in abusive relationships as adults. Definition o f Terms The use o f the word “wife” in this study does not imply any legal status o f the abusive relationship. W omen in all types o f relationships are victim s o f wife abuse, and the term wife is used here to describe women in any relationship w ith a m ale partner, including dating, marital, and comm on-law relationships. Similarly, m en who abuse women may or may not be m arried to their victim s. W omen are abused by past or present boyfiriends, husbands, and common-law partners. The term “partner” is used to describe m en who are, or have been, intim ately involved w ith the women they abuse. M any types o f abusive behaviour comprise wife abuse, including verbal abuse, physical abuse, em otional or psychological abuse, rape, sexual assault, threats, harassment. control, financial abuse, terrorism , abuse o f pets and proper^, intimidation, and isolation (Dutton, 1995; Johnson, 1996; McCue, 1995; Straus & Celles, 1990; Yllo, 1993). The women in this study have experienced many o f these abusive behaviours, and are deem ed by them selves and their counsellors to be victims o f wife abuse. Children’s exposure to wife abuse does not always involve direct observation o f violent incidents. It may consist o f overhearing a physical confix)ntation or seeing the signs o f violence on their mothers’ bodies. Whether or not children actually see a man abuse their mother, they are exposed to w ife abuse because they live in an environment where arguments, threats, and physical abuse occur. In fam ilies where wife abuse occurs, the home environm ent is strongly afiected. Children live in constant fear and apprehension about when the next violent incident will occur. They are also exposed to parental models whose behaviour is strongly affected by the differential power each person holds. Due to the pervasive negative home environment these children often live in, researchers have described children’s exposure to wife abuse as a form o f psychological m altreatm ent or emotional abuse (Bam ett et al., 1997). Brassard, Hart, and Hardy (1991) believe children are subjected to a form o f psychological maltreatm ent called “terrorizing” when they are exposed to violence or threats directed tow ard family members (p. 256). Another category o f psychological maltreatment, called “exploiting and corrupting” includes the modelling o f antisocial acts and unrealistic roles, and encouraging or condoning “deviant standards or beliefs” (Brassard et al., 1997, p. 256). Exposure to wife abuse, in my opinion, falls into both o f these categories, as children exposed to wife abuse are repeatedly exposed to verbal and physical aggression, threats o f violence against their mothers, models o f violent behaviour, unrealistic sex-roles, and the acceptance o f beliefs which condone violence against women. The Bem Sex Role Inventory (BSRI) (See Appendix A) classifies the item s on its scales as m asculine or feminine if they are considered more socially desirable for either men or women. Those items on the scales considered ‘feminine” were judged by both m en and women to be more socially desirable for a woman than a man; items considered “m asculine” were those judged more socially desirable for a man than a woman (Bem, 1981). Respondents are considered masculine when they scored high on the masculine scale and low on the fem inine scale and considered fem inine when they scored high on the fem inine scale and low on the masculine scale. Although the BSRI does not label masculine or feminine sex-typed individuals as “traditional”, I use the term “traditional” in my description o f participants if they score high on one sex-role dimension and low on the other because I believe that endorsem ent o f one type o f sex-role-specific behaviour at the expense o f the other is “traditional” behaviour in that it fits w ith society’s historical views o f what comprises appropriate “male” or “fem ale” behaviour. Basow (1992) believes traditional sex-roles and sex-role stereotypes are not based on actual differences between the sexes, but on a differential power relationship betw een m en and women. These traditional sex-role beliefs have the power to lim it what m en and women are able to do, and have a negative effect on both the individual and society (Basow, 1992). Fam ilies characterized by wife abuse often hold pervasive traditional sex-role beliefs which, if adopted by their children, may affect their relationships with others. The intergenerational transmission o f violence theory argues for “the propensity for exposure to aggression in one generation to increase the likelihood o f aggressive behaviour in a later generation” (Doumas, M argolin, & John, 1994, p. 158). This theory analyzes the role o f gender beliefs in the transm ission o f wife abuse, as individuals who endorse traditional gender beliefs are more likely to be either perpetrators or victim s o f wife abuse (Bam ett et al., 1997; Choice et al., 1995; Dutton, Starzomski, & Ryan, 1996; Celles & Cornell, 1990; Health & W elfare Canada, 1992; Kashani, Daniel, Dandoy & Holcomb, 1992; Moore, Pepler, W einberg, Hammond, Waddell, & W eiser, 1990; Osofsky, 1995). Scope and Lim itations o f Study The current study is limited to an analysis o f the sex-role beliefs o f children exposed to wife abuse in one city. Despite the small local sam ple, this study may present valuable inform ation on children exposed to wife abuse, due to the fact that regardless o f where they live, children exposed to wife abuse experience sim ilar fears and threats, and are exposed to models o f violence, abuse, control, and power. Outline o f Thesis Chapter two consists o f an introduction to w ife abuse theory, including prevalence rates, theoretical causes, sex-role beliefs o f abusers and victim s, and the effects o f children’s exposure to incidents o f wife abuse. I include such a thorough discussion o f wife abuse in order to dem onstrate the attitudes and behaviours com m on to fam ilies in which wife abuse occurs and to w hich the children in these families are exposed. I then focus on the issue o f sex-role beliefs o f children exposed to wife abuse, and in chapter three describe m y study o f the sex-role beliefs o f a group o f children exposed to wife abuse. I statistically analyze the results o f my study in chapter four, and discuss the implications o f my findings in chapter five, including possible social consequences o f children’s incorporation o f traditional sexrole beliefs. CHAPTER TWO: LITERATURE REVIEW Wife Abuse W ife abuse is one o f today’s m ost serious and widespread social problem s, affecting people o f all socioeconomic statuses, cultures, ages, and education levels. Estim ates o f wife abuse vary considerably. Health and W elfare Canada (1992) estimates that one in ten Canadian women is the victim o f abuse by a male partner. Statistics Canada’s 1993 study found that 29 percent o f married women or those who have lived in a comm on-law relationship have been physically or sexually assaulted by their partners at least once (W omen in Canada, 1995). This num ber does not include girlAiends or divorced women who have been assaulted, nor does it include those who did not report incidents o f abuse against them. Nonetheless, this 29 percent represents at least one m illion Canadian women (Johnson, 1996). Although wife abuse occurs in all types o f relationships and among all types o f people. Statistics Canada’s Violence A gainst Women survey found numerous variables to be correlated with wife abuse. Despite conflicting evidence in the wife abuse literature regarding these variables, studies have found mild to moderate relationships between wife abuse and age, education, income, type o f relationship, and the use o f alcohol (Bam ett et al., 1997; D utton et al., 1996; Johnson, 1996; McCloskey, 1996). Young wom en between the ages o f 18 and 24 appear to be more at risk o f being victims o f wife abuse, as do those in common-law relationships (Johnson, 1996; McCloskey, 1996). Couples o f low socioeconomic status and education levels are somewhat more likely to be involved in abusive relationships, although income and education appear to be less o f a factor than age or type o f relationship (Johnson, 1996; M cCloskey, 1996). In addition, m en who drink alcohol are m ore likely to assault their wives, and are m uch more likely to use severe violence against them (Celles & Cornell, 1990; Johnson, 1996). Readers may refer to B am ett et al. (1997), Celles & Cornell (1990), Johnson (1996), o r McCloskey (1996) for a thorough analysis o f factors associated w ith wife abuse. Straus and Celles (1990) have conducted extensive research on fam ily violence using the Conflict Tactics Scales (See Appendix A), finding a high incidence o f both husband and wife abuse. For some people, the occurrence o f husband abuse calls into question the greater focus on wife abuse. Despite the discovery o f approximately equal numbers o f husband and wife abuse, women’s abuse o f m en is often in self-defense (Cantos, Neidig, & O ’Leary, 1994; Straus & Celles, 1990). Furthermore, m en’s greater size and strength pu t wom en at increased risk o f injury and hospitalization (Cantos et al., 1994). Studies consistently indicate that female victims are three tim es m ore likely than males to require m edical attention for injuries sustained in spousal assaults (Cantos et al., 1994; Straus & Celles, 1990). Violence by women against their m ale partners normally would not, and could not, have the same effect. The power differences between m en and women in society also function to put women in a vulnerable position regarding abuse. Relative to men, women rem ain financially disadvantaged, and are often more dependent on their partners for economic support, particularly when they have children. Unlike women, men can most often use violence w ithout fear o f physical retaliation o r economic repercussions (Johnson, 1996). 10 Thus, although husband abuse is a problem for a m inority o f men, I chose to focus solely on wife abuse, as women are victim s o f spousal assault in far greater num bers th an men. Theoretical Analyses o f Wife Abuse Due to the emotional and political nature o f this social problem, there is significant controversy and disagreement about possible causes o f wife abuse. Num erous theories attem pt to explain wife abuse, but no one theory is universally accepted as being the most fitting. The m ost comprehensive theories combine individual, social, and cultural factors in their analyses o f wife abuse. The m ain theoretical approaches to describing and understanding wife abuse may be categorized as the psychological, the sociological, and the fem inist approaches. Another recent theory is Dutton’s analysis o f borderline p erso n alia orientation and its relationship to wife abuse (Dutton et al., 1996). Psvchological Theorv The psychological perspective on wife abuse focuses on individual personality traits o f the abuser as being responsible for the violence. Psychological disorders and m ental illness are blamed for the actions o f the perpetrator, who is labelled psychotic, paranoid, or sociopathic (Johnson, 1996). Abused women are then labelled masochists for staying w ith their partners (O’Leary, 1993). These psychiatric labels serve to decrease the responsibility o f the perpetrator for his actions by placing it on the victim (O’Leary, 1993). The value o f the psychological approach to wife abuse lies in its analysis o f the continuum o f physical aggression. Psychological studies have found that as the severity o f 11 violence increases, so too does the likelihood o f the perpetrator’s having some type o f personality disorder (O’Leary, 1993). Though this finding may help account for abuse perpetrated by men w ith specific personality traits or disorders, it does not explain why men w ithout such characteristics abuse their wives. Psychological theory fails to account for the prevalence o f wife abuse throughout society, and does not acknowledge the interplay o f individual, social, and cultural factors in the etiology o f wife abuse. Sociological Theorv W hile proponents o f the psychological theory focus on individual responsibility, sociologists focus on the influence o f society, assigning blame to a w orld which allows and essentially condones violence against women. Sociologists argue that people do not act independently o f their surroundings; they see people and their behaviour as influenced by aspects o f their social environments including age, sex, socioeconomic status, race, and ethnicity (Celles, 1993). Sociologists attem pt to address the shortcomings o f psychological theory by focusing on the structure o f the family as a powerful influence on the occurrence o f wife abuse (Celles, 1993). Sociological theory attempts to integrate num erous family characteristics with social influences that they believe make the fam ily prone to violence. Systems theory is the prim ary sociological theory o f family violence. It outlines various family characteristics that put the family at risk o f family violence including; a) a large amount o f time spent together; b) family involvement in a wide range o f activities and interests; c) intensity o f involvement; d) impinging activities; e) beliefs in the right to 12 influence family members’ values, attitudes, and behaviours; Q potential for conflicts between generations and sexes due to age and sex differences; g) assignm ent o f roles and responsibilities based on age and sex rather than interest or competence; h) fam ily privacy and isolation from society; i) personal, social, m aterial, and legal commitment to the family; j) susceptibility to stress through fam ily changes and transitions; and k) intim acy and em otional involvement (Gelles, 1993, p. 36). Fam ily systems theorists believe that the origins o f the problem o f violence lie in the nature o f the family, not specifically in the relationships between husband and wife (Kurz, 1993). The concepts o f positive and negative feedback are integral to the systems theory o f wife abuse. Positive feedback, such as the woman trying to increase her power in the relationship, cause change in the fam ily system (Becvar & Becvar, 1996). Negative feedback, such as the woman’s staying w ith her abusive partner, fimctions to m aintain the abusive family system. (Becvar & Becvar, 1996). Systems theory fails to recognize the widespread social legitim ization o f m en’s control and violence against women, and does not account for the fact that the m ajority o f spousal assault consists o f m en’s abuse o f their wives. If the belief o f system s theorists that fam ilies are generally susceptible to violence against each other were true, one would assume that sim ilar numbers o f men and women would be the victim s o f abuse. However, women are overwhelmingly the victim s o f fam ily violence, due at least in part to the imequal power o f husbands and wives (Kurz, 1993; Yllo, 1993). Fem inist theorists criticize systems theory for its suggestion that violence betw een fam ily members is a m atter o f conflict o f interest rather than one o f male power and 13 dom ination (Kurz, 1993; Yllo, 1993). They argue that fam ily violence is a “tactic o f entitlem ent and power that is deeply gendered, rather than a conflict tactic that is personal and gender-neutral” (Yllo, 1993, p. 57). Systems theory does not incorporate social realities regarding wom en’s lack o f power relative to that o f men. Socialization o f gender theories, particularly gender schema theory, attem pt to address this issue by studying m en’s and w om en’s differential status in relationships. Socialization o f Gender Theories Developm ental psychologists view the childhood socialization process as one in w hich parents teach or transm it rules and expectations to their children (Jacklin & Reynolds, 1993). Social learning theory and gender schema theory are the two dominant theories in this area. Social learning theorv For m any years, social learning theory has dom inated children’s socialization research. Its analysis o f m odelling forms a succinct and comprehensive theory describing the process by w hich children incorporate many o f their parents’ values, attitudes, and behaviours. Children are known to imitate the behaviour o f others, especially others who are sim ilar to them in some way. Boys will imitate their fathers, while girls pattern their behaviour after their mothers. Children are more likely to im itate the same forms o f violence they are exposed to (Choice et al., 1995). If children are exposed to role models o f m ale violence and female victim ization and repeat the behaviours typical o f abusers or victim s, there m ay be predictable and frightening consequences. 14 Children exposed to wife abuse develop either passive or aggressive problem-solving strategies (Copping, 1996; Tutty & W agar, 1994). Œ ven the tendency for children to imitate sim ilar or same-sex models, female children may be more likely to become passive in situations o f conflict, and boys more likely to become aggressive. Studies have found a positive relationship between men’s exposure to their fathers’ violence and the tendency to use violence against their own wives (Barnett et al., 1997; Choice et al., 1995; Dutton et al., 1996; Celles & Cornell, 1990; Health & Welfare Canada, 1992; Kashani et al., 1992; M oote et al., 1990; Osofsky, 1995). Social learning theory attem pts to explain this apparent intergenerational transmission o f violence, suggesting that children exposed to wife abuse learn that violence is an acceptable way to deal w ith conflict. Despite the consensus among researchers that violence is transm itted intergenerationally, social learning theory does not, in itself, adequately account for the prevalence o f wife abuse throughout society. Specifically, the social acceptance o f wife abuse and the perpetration and victim ization o f individuals who were not exposed to wife abuse as children are not addressed. In addition, many children exposed to wife abuse are not involved in abusive relationships as adults, despite the powerfiil influence o f parental role m odels. Gender schema theorv Gender schema theory attempts to address the shortcomings o f social learning theory by explaining that children are not passive imitators o f abusive behaviours. Rather, proponents o f gender schema theory see children as using a process o f “selective cognition” 15 in their formation o f gender schemas (Jacklin & Reynolds, 1993, p. 200). A schema is a set o f ideas that an individual uses to organize and filter information (Jacklin & Reynolds, 1993). Children develop gender schemas based on the information they receive relating to gender, allowing them to sort people, behaviour, and attributes into society’s definitions of m asculinity and femininity (Jacklin & Reynolds, 1993). Children exposed to wife abuse may therefore develop gender schemas which pair violence with m asculinity and victimization w ith femininity. Children’s “male” category w ill thus reflect the greater power men have in fam ilies and in society (Jacklin & Reynolds, 1993). Gender schema theory incorporates the role o f gender in social learning but does not sufficiently address the consequences o f patriarchy for men and women as thoroughly as fem inist theory does. Yllo (1993) does not see distinctions between m en and women as inherent or fimctional; they are social constructs which create and m aintain male power w ithin the family. The social constructions o f masculinity and fem ininity which serve to increase the power o f m en at the expense o f wom en need to be altered to create a more equal and respectful balance o f power in the family, which in turn would decrease the likelihood o f wife abuse occurring. F em inist Theorv In recent years, fem inist theory has become the dom inant theoretical model in the study o f wife abuse (Celles & Loseke, 1993). The feminist perspective challenges the psychological and sociological perspectives by moving beyond individual and social problem s associated w ith wife abuse to focus on the effects o f gender socialization and 16 patriarchy on m en and women. Fem inist theory lends itself well to integration w ith other theories, combining several components o f social learning theory and gender schem a theory to create a comprehensive theory o f the history, causes, and effects o f wife abuse. Fem inist theorists agree w ith social learning theorists and gender schema theorists that m en and wom en are socialized to develop sex-typed beliefs and attitudes. However, fem inists attribute the prevalence o f such beliefs and attitudes to social laws and practices that im plicitly and explicitly approve o f m ales’ greater power (Johnson, 1996). Fem inists argue th at wife abuse cannot be adequately understood unless gender and power are taken into account (Yllo, 1993). According to feminists, the patriarchal social system is responsible for both m en’s and wom en’s gender-role socialization. Patriarchy refers to social structures that enable men to feel entitled to power and control in their relationships (Johnson, 1996; M cCue, 1995; Sm ith, 1990). Feminists believe society views m en as the dominant class, with wom en placed in a secondary and inferior position (McCue, 1995). In this view, society defines men as “dominant, strong, authoritarian, and aggressive”, while women are traditionally viewed as “dependent, passive, and submissive” (M cCue, 1995, p. 13). Feminists consider social acceptance and the condoning o f male superiority and aggression to be solely responsible for violence against women. They see women’s victim ization as a social problem based on the psychological control and physical dom ination o f women by men, and believe w ife abuse can only be elim inated when women and m en are truly equal (Yllo, 1993). Proponents o f the fem inist theory believe male violence against women exists to such an extent due to society’s and the fam ily’s view o f m en as having higher status and more 17 pow er and authority than women (Johnson, 1996). Fam ilies are a£fected by patriarchy in that they often “embody traditions, roles, and beliefs about the proper place for men and women and thus provide(s) both the structure and an ideology that endorses a higher status role for men” (Johnson, 1996, p. 158). W ife abuse occurs as a natural result o f this unequal relationship between m en and women. Fem inists argue that early sex-role socialization conditions girls to become subm issive victim s, while boys leam to act as perpetrators o f violence (M cCue, 1995). They leam that “violence is the basis o f power and control in fam ilies, that women have fewer rights and less value than men, and that fathers have a right to use violence against their wives” (Johnson, 1996, p. 172). Fem inist theory has found much support through em pirical and conceptual research. Gender inequality may explain variations in the incidence and rates o f wife abuse (Celles, 1993). However, psychological, sociological, and biopsychosocial theorists find its focus on patriarchy lim ited (Dutton et al., 1996; Celles, 1993; O’Leary, 1993). These theorists believe fem inist theory focuses on gender and patriarchy at the expense o f other im portant aspects o f wife abuse. Fem inists them selves agree with some o f the criticism o f their theory, as they recognize that no one theory can adequately explain why only some m en abuse their wives (Yllo, 1993). Fem inists support continued research into the many factors associated with wife abuse, such as low income and education, stress, alcohol use, and childhood exposure to wife abuse (Yllo, 1993). 18 Abusive Personality Theorv Dutton’s recent theory correlating borderline personality orientation with wife abuse attempts to integrate biological, psychological, and social characteristics o f abusive men (Dutton et al., 1996). The “abusive personality” theory suggests that a combination o f characteristics and life experiences make certain men more likely to become abusive (Dutton et al., 1996). Specifically, his study found that abusive m en scored significantly higher than a control group o f non-abusive men on measures o f childhood exposure to wife abuse, abusive behaviour, and abusive personalia (Dutton et al., 1996). Past research has often focused on only one variable, such as psychological disorders, the social environment, fam ily characteristics, modelling, or gender beliefs in its analysis o f wife abuse. The im plications o f Dutton’s research are that fam ily violence researchers need to consider the effects o f m ore than one variable on men’s likelihood o f being abusive and develop methods o f assessing various characteristics and experiences o f abusive men. Summary o f Theoretical Analyses o f W ife Abuse Although theories o f fam ily violence in general differ w ith regard to their conceptual focus, all support the integration o f different aspects o f their theories. Psychological theorists recognize the lim its o f psychological characteristics as the cause o f wife abuse, as most men who abuse their wives do not have any psychological disorder. However, they also recognize that psychological factors are often involved in cases o f severe violence. Sociological theorists see the strength o f their perspective as the understanding o f fam ily characteristics which make fam ilies vulnerable to violence. Social learning theory builds on sociological 19 theory by considering the process through which children incorporate attitudes, beliefs, and behaviours o f their parents, and gender schema theory takes this m odel further by analyzing gender effects in greater detail. Fem inist theory moves beyond individual and fam ilial beliefs about gender to focus on the effects o f patriarchy for both m en and women. Fem inists do not see sexism as one factor in the etiology o f wife abuse; they believe it is an encompassing feature o f the phenomenon o f violence against women. Dutton’s argum ent for the existence o f an abusive personalia brings together biological, psychological, and experiential factors to account for the prevalence o f wife abuse, and shows m uch prom ise for use in hiture wife abuse research. Taken together, these theories illustrate the complexity o f the issue o f wife abuse. M any variables are involved in its creation, maintenance, and elim ination. Studying children exposed to wife abuse is one way to develop our understanding o f this social problem , as each theory identifies the increased likelihood o f children who are exposed to wife abuse becoming involved in abusive relationships later in life. Theoretical Perspective o f Current Study Despite the distinct theoretical focus o f each o f the above theories o f wife abuse, they are united in their belief that childhood exposure to wife abuse is a risk m arker for involvem ent in abusive relationships as adults, as either the perpetrator or victim o f violence. There are many paths by w hich children exposed to wife abuse may incorporate beliefs condoning violence in relationships. Systems theorists argue that the fam ily system is vulnerable to abuse due to its inherent organizational and structural characteristics. 20 Socialization o f gender theories agree that the fam ily has a powerful influence over children’s behaviour, but focus specifically on m odelling and gender schemas to explain w hy violent couples are m ost often com prised o f m ale abusers and female victim s. Fem inist theory expands the analysis o f gender in its theory o f social and fam ilial patriarchy, arguing that boys exposed to wife abuse may leam attitudes common to perpetrators o f violence and girls exposed to wife abuse may leam those common to victims. Dutton e t al. (1996) found boys’ childhood exposure to wife abuse to be correlated with abusing their w ives in adulthood. These theories lead directly to m y research questions by suggesting that childhood exposure to wife abuse is a critical factor in the development o f abusive adult relationships. However, they do not identify the specific sex role beliefs and attitudes w hich m ay make children exposed to w ife abuse vulnerable to involvement in abusive relationships as adults. My study integrates concepts o f fam ily systems theory, social learning theory, gender schema theory, fem inist theory, and abusive personalify theory in an attem pt to analyze possible relationships betw een traditional sex-role beliefs and children’s exposure to wife abuse. Sex-Role Beliefs o f M ale Abusers and Female Victims Researchers are often interested in personality characteristics o f abusers and their victims. However, because wife abuse exists in every facet o f society, there is m uch variation in such characteristics. Abusive men and their female victim s are not easily recognized or identified. They come from all educational and econom ic levels, races, religions, and backgrounds, lending further credibility to the fem inist idea that m en’s 21 violence against women is taught, developed, and practised throughout society (McCue, 1995). One characteristic often present in families characterized by wife abuse is a belief in traditional sex roles. Beliefs about the rights o f husbands to assert control over wives form a component o f battering relationships (Johnson, 1996). Men who abuse their wives often feel entitled to control and dominate their partners. Traditional, sex-typed beliefs reflect themes o f power and control which are common in abusive relationships. Extensive studies have been conducted on types o f male batterers, and despite the great variety o f characteristics within this group, it is united by a belief in male superiority. As early as 1979, Walker identified a belief in male superiority as a trait typical o f abusive men. Since then, other researchers have substantiated this finding. DeKeserdy and Kelly (1993) conducted a study o f abusive m en’s beliefs and attitudes toward women. Specifically, they questioned whether men who believe they have the right to dominate women in relationships have higher rates o f wife abuse than those who hold more egalitarian beliefs. Results fi’om the sample o f 1307 male college and university students showed that although m ost students did not believe in male dominance, those who did were m ost likely to physically assault their partners (DeKeserdy & Kelly, 1993). Hurley and Jafte (1990) found that violent families are typified by an unequal power relationship between husband and wife. They identified strong patriarchal influences in the family, which function to increase the power o f the male perpetrator. Michael Smith, a Canadian sociologist, conducted a study using a random sample o f women, asking them to consider their husbands’ beliefs regarding their right to control and 22 dominate women in intim ate relationships (1990). He found support for the hypothesis that men who endorse their right to dominate women would have higher rates o f wife abuse. The results firom the 600 fem ale respondents showed that m en who hold traditional beliefs and condone violence against women in the fam ily are more likely to behave violently toward m arital partners than m en who hold more egalitarian beliefs. Smith’s study found both traditional sex-role beliefs and attitudes approving o f violence against women to be statistically significant predictors o f whether a man ever abused his wife. The stronger those beliefs, the greater the probability^ that the wife had been beaten. Despite the demonstrated relationship between traditional beliefs and the likelihood o f m en abusing their wives, it is important to note that a m uch larger proportion o f men hold traditional beliefs (18-53%) than abuse their wives (3-20%) (Johnson, 1996, p. 160). The identification o f traditional sex-role beliefs as a common denominator in the personality characteristics o f male abusers is not enough to say that it causes wife abuse. Estimates o f Children Exposed to W ife Abuse Estimates o f children exposed to wife abuse are usually based on parents’ reports, especially those o f the mothers. Studies have shown that parents often underestimate the extent o f their children’s exposure, perhaps due to an unwillingness to consider the harmful effects o f their behaviour on their children (Sternberg, Lamb, Greenbaum, Cicchetti, Dawud, Cortes, Krispin, & Lorey, 1993). Parents may assume that their children are unaware o f the abuse, particularly if the incidents occur while the children are believed to be sleeping, or while they are in another room. However, interviews w ith children o f abused women have 23 found that alm o st all can describe incidents o f wife abuse that their parents did not know they had been exposed to (Jaffe, W olfe, & W ilson, 1990). A recent study o f fathers’, m others’, and children’s reports o f children’s exposure to wife abuse found that although m others’ and fathers’ reports were sim ilar, parents’ and children’s were quite different (O’Brien, John, M argolin, & Erel, 1994). Apparently, estim ates o f how m uch wife abuse children are exposed to depends on who is asked. A national survey o f abused women residing in shelters found that 25 percent thought their children had been exposed to their abuse (Tomkins, Mohamed, Steinman, M acolini, Kenning, & Afrank, 1994). In a study o f m arried women who were victim s o f wife abuse, 39 percent said their children had been witnesses (Johnson, 1996). W olfe and Jaffe (1991) found that children observed 68% o f wife assaults in which charges were laid and Health and W elfare Canada (1992) estim ates that children are exposed to as many as 80% o f all incidents o f wife abuse. Jaffe, W olfe, and W ilson (1990) believe that due to underreporting o f wife abuse, 3.3 m illion is a conservative estim ate o f the number o f Canadian children exposed to wife abuse each year. These numbers reflect the severity and potential im pact o f wife abuse, not only for the direct victim s o f the physical violence, but for their children as well. Effects o f Children’s Exposure to W ife Abuse The effects o f wife abuse are often apparent on the female victims; physical bruises and injuries, emotional distress, and feelings o f hopelessness are common. Recent studies o f children exposed to wife abuse describe a w ide range o f problems, w ith some children seem ingly unaffected, and others displaying clinical levels o f behavioural, physical, and 24 psychological problem s. Exposure to wife abuse affects children’s health, thoughts, feelings, and actions in a num ber o f negative ways. Table 1 outlines the findings o f several recent studies o f children exposed to wife abuse. Table 1: Behavioural. Phvsical. and Psvcholoeical Effects o f Exposure to W ife Abuse Effect Behavioural Effects: sl) internalizing: anxiety withdrawal passivity bl externalizing: aggression Reference Hughes, Parkinson, & Vargo, 1989; Hurley & Jaffe, 1990; Kashani et al., 1992; Moore et al., 1990; Osofsky, 1995; Tutty & Wagar, 1994 Moore et al., 1990; Tapp & Hinish, 1992 Hurley & Jaffe, 1990; Suh & Abel, 1990; Tapp & Hinish, 1992 Fantuzzo e ta l., 1991; Hurley & Jaffe, 1990; Kashani et al., 1992; M oore et al., 1990; Suh & Abel, 1990; Tapp & Hinish, 1992 im pulsivity Hurley & Jaffe, 1990; Tapp & H inish, 1992 delinquency Fantuzzo et al., 1991; Moore et al., 1990 Phvsical effects: somatic problem s Moore et al., 1990; O ’Keefe, 1994; Tutty & Wagar, 1994 poor sleep habits Kashani et al., 1992; Osofsky, 1995; Tapp & Hinish, 1992 enuresis Fantuzzo et al., 1991 nightm ares Fantuzzo et al., 1991; Kashani et al., 1992; Tapp & Hinish, 1992 stomach aches, headaches, ulcers McCue, 1995; Tapp & Hinish, 1992 Psvcholoeical effects: depression Fantuzzo et al., 1991; Moore et al., 1990; O ’Keefe, 25 1994; Sternberg et al., 1993; Tutty & Wagar, 1994 irritability Kashani et al., 1992 decreased attention and concentration, intrusive thoughts Osofsky, 1995 low self-esteem Fantuzzo et al., 1991 ; Moore et al., 1990; Tutty & Wagar, 1994 powerlessness Moore et al., 1990 limited em pathy and poor prosocial competence Fantuzzo et al., 1991; Hurley & Jaffe, 1990; Osofsky, 1995 constricted and inhibited emotions Hurley & Jaffe, 1990;Kashani et al., 1992 post-traum atic stress disorder, negative self-im age_____________ Osofsky, 1995___________________________________ Gender Differences in Behavioural Effects Behavioural effects o f exposure to wife abuse fall into two categories: a) internalizing behaviours, and b) externalizing behaviours. Many studies find boys more likely to display externalizing symptoms than girls, while others do not reveal gender effects related to exposure to w ife abuse. Jaffe, W ilson, & Wolfe (1988) and Jaffe, Wolfe, and W ilson (1990) found that boys exposed to wife abuse displayed problem behaviours sim ilar to boys who had been abused them selves, becoming aggressive, disobedient, and destructive. W hile boys often express the effects o f exposure to wife abuse openly, girls may become passive or withdrawn (Jaffe, Wolfe, & W ilson, 1990). Girls express the traum a o f being exposed to their m others’ abuse differently than boys, becoming “passive and withdrawn as they witness the assaults and see that their mothers are powerless to stop them ” 26 (Johnson, 1996, p. 172). However, the invisible effects on girls may eventually express them selves externally during adolescence in the form o f aggression, rebellion, and high-risk behaviour (Henning, Leitenberg, Covery, Turner, & Bennett, 1996 ; Tapp & H inish, 1992). Im m ediately following an incident o f abuse, young girls may become withdrawn, but they often show aggressive and impulsive behaviour later in life, particularly in adolescence. O’Keefe (1994) found boys and girls to be equally at risk for externalizing and internalizing problems, demonstrating that boys do not necessarily react aggressively and girls passively. Regardless o f who reacts aggressively and who reacts passively, both internalizing and externalizing behaviours lead to increased difSculties for these children, as aggression and im pulsivity m ay isolate them 6om peers, and withdrawal and passivity can have a powerfiil effect on self-esteem, problem -solving abilities, and the ability to express feelings. M any factors are involved in the interplay among children’s health, thoughts, feelings, and behaviours. A fier continued exposure to wife abuse, children may believe they are powerless. They may soon begin to act as such, becoming either withdrawn and passive or aggressive and impulsive. These behaviours increase the risk o f health problem s and may lead to poor relationship development, w hich has continued long-term consequences for them as they grow up. Sex-Role Beliefs and the Intergenerational Transmission o f Violence Theory The intergenerational transmission o f violence theory is often debated in contemporary research, especially as it applies to wife abuse. Like any other theory o f wife 27 abuse, it does not sufficiently explain and account for the prevalence o f wife abuse in society and throughout fam ilial generations. Nonetheless, research in the field o f wife abuse consistently finds a greater fi^quency o f wife abuse among adults who were exposed to wife abuse as children (Celles & Cornell, 1990; H urley & Jaffe, 1990; Jaffe, Hurley, & W olfe, 1990; Jaffe, W ilson, & W olfe, 1988). M en exposed to wife abuse as children dem onstrate greater proclivity for perpetrating wife abuse, w hile women exposed to wife abuse as children are more likely to be victimized. The presence o f traditional sex-role beliefs is one factor which contributes to the negative cycle o f wife abuse, as children adapt and incorporate their parents’ sex-role beliefs (Brassard et al., 1991; Health & W elfare Canada, 1992; Hurley & Jaffe, 1990; Kashani et al., 1992; M oore et al., 1990; Osofsky, 1995; Tapp & Hinish, 1992). From an early age, boys are commonly taught to be tough, and girls to be submissive. These lessons are not solely taught in the fam ily; children are bombarded with sex-role stereotypes on television, in m ovies, in literature, and in sports. However, children exposed to wife abuse in their homes are ofien exposed to the extreme forms o f these stereotypes. Boys are exposed to m odels o f power, dominance, and control, and girls to models o f passivity, subordination, and vulnerability. The direct modelling o f such behaviours in the fam ily is a powerful force w hich adds to the daily exposure to and influence o f other types o f sex-role stereotyping. Growing up in male-dominated fam ilies where wife abuse occurs affects children and their understanding o f gender roles. If the fam ily has traditional sex-role beliefs and expectations, girls are trained to pattern their behaviour after their m others, and boys are taught to expect the same authority and privileges as their fathers (Tapp & Hinish, 1992). 28 Hurley and Jaffe (1990) found that children assim ilate attitudes and values that perpetuate the cycle o f violence in the family, particularly those that foster and condone aggression toward women. W hen attitudes o f an individual are s lu ^ d that support the use o f physical force against a woman, physical aggression is much more likely (O’Leary, 1993). W hile many men and women do not endorse violence against women, others’ attitudes are more strongly influenced by a society which suggests that violence against wives is acceptable (O’Leary, 1993). Children exposed to wife abuse leam that violence is an ^p ro p riate way to resolve conflict and that m en’s violence toward women can be rationalized and accepted (Health & W elfare Canada, 1992; Hurley & Jaffe, 1990; Kashani et al., 1992; Moore et al., 1990; Osofsky, 1995). They are therefore more likely to use violence themselves, and internalize these lessons about conflict, power, control, and the differential value and privileges o f the genders (Tapp & Hinish, 1992). There are serious consequences to children’s exposure to their parents’ sex-role beliefs and attitudes about violence in relationships. Numerous researchers have foimd that children identify w ith their parents based on gender and will use their parents’ relationship as a model for their own future relationships (Groves, Zuckerman, Marans, & Cohen, 1993; Hurley & Jaffe, 1990; Jaffe, Hurley, & Wolfe, 1990). When children are exposed to the abuse o f their mothers, they may rationalize the abuse, believing, “the man is boss,” “she provoked it,” and “you have to put up with it” (Hurley & Jaffe, 1990, p. 472). Jaffe, Hurley, and W olfe (1990) suggest that when boys identify w ith a violent father and girls identify w ith an abused m other, they may develop attitudes and behaviours common to perpetrators and victim s o f violence, respectively. When parents deal with conflict through aggression or 29 withdrawal, boys may leam to deal w ith relationship conflict through aggressive m eans, and girls through passive means (Tutty & Wagar, 1994). This places children exposed to wife abuse at risk for responding with violence or being victim ized in adult relationships (T u t^ & Wagar, 1994). Perhaps due to a lack o f appropriate role models, these children do not develop conflict resolution skills, nor do they have the ability to avoid the use o f violence and aggression during a conflict (Johnson, 1996; M oore et al., 1990). Therefore, they m ay be more likely to become involved in abusive relationships in their teenage or young adult years, setting up a pattern o f relating to partners that involves the use o f intimidation, control, and violence by perpetrators and eventual submission and p assiv i^ by victims. Studies have found a relationship between childhood exposure to violence and feture involvement in abusive relationships (Barnett et al., 1997; Choice et al., 1995; Celles & Cornell, 1990; Johnson, 1996). Boys exposed to wife abuse may carry lessons about gender and power into adulthood and perpetuate the cycle o f violence by abusing their own wives (Health & W elfare Canada, 1992). M en exposed to their m others’ abuse were up to three tim es more likely to abuse their own wives compared to men who grew up in non-violent homes. The Statistics Canada 1993 Violence A gainst Women Survey found that in its telephone sample o f 12,300 women, 36 percent o f the women’s abusive male partners had been exposed to wife abuse as children, compared to 12 percent o f m en who had not been exposed to w ife abuse as children (Johnson, 1996). Boys exposed to the abuse o f their mothers leam the attitudes and behaviours typical o f abusive men, which make them 30 significantly m ore likely to become abusive as adults (Johnson, 1996; Straus & Celles, 1990). Girls are powerfiilly affected by childhood exposure to their m others’ abuse, as they see their fathers’ assaults and their mothers’ apparent inability to stop the abuse. These experiences m ay teach girls that abuse is something they must endure. If violence occurs in their adult relationships, girls m ay feel powerless to stop it (Health & W elfare Canada, 1992; Henning et al., 1996). The Canadian Violence Against W omen Survey found that women exposed to wife abuse as children experienced abuse by their husbands at a rate alm ost twice as high as women who grew up w ithout such exposure (Johnson, 1996). The pattern was also strong in the women’s previous battering relationships, where 67 percent o f women exposed to wife abuse as children experienced it as adults, compared to 43 percent o f women not exposed to w ife abuse in childhood (Johnson, 1996). Neither gender is immune to the consequences o f witnessing wife abuse. The above evidence provides support for the intergenerational transm ission o f violence theory. However, caution m ust be taken before accepting it as a causal explanation for this phenomenon. M any men exposed to wife abuse as children do not grow up to abuse their wives, and many abusive men were never exposed to wife abuse as children (Johnson, 1996; O’Keefe, 1994). W hile there is a relationship between childhood exposure to wife abuse and future involvem ent in abusive relationships, the lack o f a perfect association suggests that there are other factors involved. Thus, while there is em pirical support for the theory that violence is transm itted from one generation to another through m odelling and the 31 incorporation o f traditional sex-role beliefs, other factors may intervene to break or start the cycle. O ’Keefe (1994) believes that learning to perpetuate or endure wife abuse is a developm ental and interactive process that involves more than modelling certain parental behaviours. Exposure to wife abuse is not the only factor which increases the risk for violence in intimate relationships, but it is an important one (Gelles & Cornell, 1990; Moore et al., 1990). Rationale for Present Study The present study attem pts to fill a gap in the research on children exposed to wife abuse. W hile we know that children exposed to wife abuse are at risk o f numerous behavioural, physical, and psychological problems, our knowledge o f the effects o f exposure to wife abuse on their sex-role beliefs rem ains limited. Although several studies suggest that a belief in traditional sex roles is a risk marker for involvem ent in abusive relationships, no studies have been conducted w hich focus specifically on the gender beliefs o f children exposed to wife abuse. Investigating this unexplored area o f research could be a critical step in our understanding o f the possible intergenerational transmission o f wife abuse. If wife abuse is transm itted intergenerationally, children are indeed the m ost appropriate focus for prevention. Since sex-role beliefs are learned rather than inherited, we have the ability to teach children more egalitarian views w hich would decrease their risk o f becom ing involved in abusive relationships as adults. Children exposed to wife abuse are by no means predetermined to be perpetrators or victim s o f wife abuse. By studying their experiences o f wife abuse and the subsequent 32 im pact on their sex-role beliefs, we can develop a better understanding o f the invisible ways in which wife abuse is harm ing children and may be setting them up for serious relationship problems as adults. Hypotheses My research question asked whether children exposed to wife abuse view themselves as more traditionally masculine or fem in in e than children not exposed to w ife abuse. I broke this question into one hypotheses for each gender, and then tested each hypothesis using scores on the masculine and feminine scales o f the Children’s Sex Role Inventory (CSRI). The first hypothesis stated that boys exposed to wife abuse view them selves as more m asculine and less feminine than boys not exposed to wife abuse. Hypothesis 1(a) compared the m asculinity scores o f the boys exposed to wife abuse (EWA) w ith those o f the boys not exposed to wife abuse (NEWA). Ho Mb-ewaCM) Mb-aewi(M) lt|>-ewi(M) ^ M’b-oewi(M) Hypothesis 1(b) compared the fem ininity scores o f the boys exposed to wife abuse with those o f the boys not exposed to wife abuse. H q- ~ l^b-aewim both Agency B and C Children 10 2* 2* 12 Exclusionary criteria for treatment group participation included: a) the child’s living in foster care, b) ongoing wife abuse in the fam ily, and c) the child’s not being betw een the ages o f 6 and 12. I spoke to the mothers about these criteria before meeting the fam ily. Therefore, no treatment group participants were excluded from the study once it began. Recruitment o f Comparison Group Participants The comparison group included 12 children who had not been exposed to any form o f abuse against their mothers. Seven o f the fam ilies volunteered to participate after learning about the study through a letter distributed to each student in grades one to seven at an elem entary school in Prince George, BC on April 30,1997 (see Appendix C). Eight teachers distributed a total o f 213 letters to students and parents later returned the completed form to the school where I collected them and contacted the women by telephone. Six women agreed to participate, while six others declined. Two weeks later, teachers distributed follow-up letters reminding parents o f the study and asking interested women to contact me directly (see Appendix C). One woman responded to this letter and agreed to participate. 37 I also posted a copy o f the original recruitm ent letter on the bulletin board at Agency A inviting fam ilies to participate as part o f the com parison group (see Appendix C). One woman contacted m e in response to this letter and agreed to participate. Exclusionary criteria for comparison group participation included: a) the child’s living in foster care, b) wife abuse in the fam ily at any point in the child’s life, and c) the child’s not being between the ages o f 6 and 12. No fam ilies who volunteered for the com parison group were excluded from the study. Independent Variable The independent variable in this study, exposure to wife abuse, was operationalized in several ways. A ll child participants from Agency A were identified by their therapists as having been exposed to acts o f violence against their mothers by a male parmer. The two child participants recruited from Agency B participated in the agency’s program for children exposed to w ife abuse and their m other participated in a support group for wom en and couples who have experienced wife abuse. These programs do not specifically address sexrole beliefs in their curricula, and therefore are not believed to have affected the children’s perceptions o f sex roles. The woman who learned o f the study through her involvem ent at Agency C was self-identified as having experienced abuse to which her two children were exposed. W hile it would be valuable to have the children describe the violence they have been exposed to, asking children to recall and describe incidents o f violence against their mothers raises num erous concerns. Based on the age o f the child participants and on the desire to 38 protect them firom possible harm or emotional traum a, I chose to rely solely on m others’ reports o f violence witnessed by their children. To assess the degree to which each child was exposed to wife abuse, the m others com pleted one adapted version o f the Conflict Tactics Scales (GTS) for each o f their children participating in the study (Straus, 1979) (see Appendix A). A ll items regarding verbal or physical aggression were prefaced w ith the question, “How many times has your partner or ex-partner done the following in front o f (child’s name) (within sight or hearing)?”. D ifferent tim e frames accompanied each item to increase the accuracy o f m others’ recollection o f specific incidents o f abuse. For exam ple, one item asked, “During a conflict w ith you, how many tim es has your partner or ex-partner insulted or sworn at you in front o f (child’s nameYl”. The woman then chose the appropriate category of: 0 ,1 , 2, 3-5, 6-10,1120, or m ore than 20 for each o f four tim e frames (the past 6 months, the past 12 m onths, the past 5 years, or ever). For the purposes o f data analysis, the women’s answers were converted to the m edian number o f each interval. Answers o f 3-5 were scored as a 4, 6-10 as 8, 11-20 as 15, and more than 20 as 25. Grych, Seid, and Fincham (1992) believe that parent reports may not provide accurate estim ates o f children’s exposure to violence. They refer to studies in which parents were found to either underestim ate or overestimate children’s awareness o f conflict between their parents (Grych e t al., 1992). Parents ofren assume that children are not aware o f conflict which occurs in another room o f the house, or w hile the children are believed to be asleep (Grych et al., 1992, p.559). In other cases, parents falsely believe that their children are aware o f m ore subtle conflict between parents (Grych et al., 1992, p.559). 39 The fam ilies participating in the treatm ent group have acknowledged that w ife abuse is an issue in their fam ilies which may have affected their children’s behaviour. These women, by participating in a study about violence against women, are taking an active step towards recognizing the effects o f wife abuse on their children, and are therefore believed to be making an honest attem pt to estim ate the extent o f abuse witnessed. In addition, a precise estim ate o f abuse witnessed is not necessary for this study, as I am studying possible relationships between the three types o f abuse witnessed (verbal abuse, mild violence, and severe violence) and each child’s score on the CSRI. I follow ed a standard procedure for collecting inform ation 6om all participants. The mothers and children in both groups were given the same instructions for completion o f their questionnaires, and all experienced the same debriefing procedure. Because it was not possible for me to be blind to each child’s condition, I maintained a standard procedure by following a general script for the children’s completion o f the CSRI (see Appendix D). Instruments The mothers in both the treatm ent and comparison group completed four questionnaires: a) a demographic questionnaire, b) the Conflict Tactics Scales (CTS) (Straus, 1979), c) an adapted version o f the CTS (Straus, 1979), and d) the Bern Sex Role Inventory (Bern, 1981) (Appendix A). The children completed one instrument, the Child Sex Role Inventory' (Boldizar, 1991) (see Appendix A). 40 Demographic Q uestionnaire The demographic questionnaire was designed to collect inform ation about each fam ily’s members, socioeconomic status, occupation, education level, and history o f abusive relationships. This inform ation was then used to compare the treatm ent and com parison groups. Conflict Tactics Scales The Conflict Tactics Scales (1979) are well-known and frequently used instrum ents for m easuring verbal and physical fam ily violence. Respondents estim ate how m any tim es various incidents o f reasoning, verbal aggression, and physical aggression have occurred in their families. Reliability values o f internal consistency range from .42 to .76 for the reasoning scale, and from .62 to .88 for the verbal violence scale (Bagarozzi & Schumm, 1989, p. 166). Straus has compiled a list o f studies which used the CTS and foimd alpha values ranging from .42 to .50 for the reasoning scale, .62 to .80 for the verbal aggression scale, and .69 to .88 for the violence scale (Straus & Gelles, 1990, p. 64). The low reliability o f the reasoning scale is largely due to the fact that it is composed o f only three items. In the present study, the reasoning items were not relevant to the hypotheses, and were therefore not analyzed. Thus, their low reliability w ill not affect the results o f this study. The conceptual focus o f the scales is the physical violence scale, w hich displays the highest reliability o f the three scales, w ith values ranging from .42 to .96 (Bagarozzi & 41 Schumm, 1989, p. 166). Only 5 o f 17 studies using the CTS have shown alpha coefficients o f less than .80 for the physical violence subscale (Bagarozzi & Schumm, 1989, p. 166). The high reliability o f the physical violence subscale is crucial, as this is the scale o f m ost importance to the present study. Concurrent validity o f the subscales has been established through correlations between individual fam ily m embers’ reports o f violence and by husbands and wives completing the scales in reference to their relationships (Bagarozzi & Schumm, 1989; Straus & Gelles, 1990, p.40). Large differences found between husbands’ and w ives’ reports o f violence reflect comm on under-reporting by perpetrators (Straus & Gelles, 1990, p. 69). In this study, the wom en’s reports o f violence provide the only estim ate o f violence witnessed by the children. Because the violence itself is not the focus o f this study, the victim ’s reports are sufficient for determ ining the extent o f wife abuse each child witnessed. Construct validity o f the CTS has been assessed by comparing the findings o f the scales to both theoretical and practical studies about fam ily violence and conflict resolution strategies. Num erous studies on topics such as the intergenerational transm ission o f violence, risk factors for fam ily violence, health problems associated with fam ily violence, and the effects o f children’s witnessing violence have concluded that the CTS assesses relationships between different variables associated w ith fam ily violence (Straus & Gelles, 1990). The CTS have withstood m uch criticism and controversy since their development. Criticism o f the CTS which pertains to this study focus on the sm all num ber o f violent acts described in the scales, the potential inaccuracy o f self-reports, and the lack o f attention to the 42 context o f violent incidents (Dobash, Dobash, W ilson, & Daly, 1992; Straus & Gelles, 1990). The sm all num ber o f violent acts is actually preferred for this study, as the wom en are being asked to list very personal and painful incidents. Expecting them to describe numerous incidents is unnecessary, as I require information on the general type, frequency, and severity o f violence the child participants were exposed to as opposed to information on the precise nature o f the women’s violent relationships. Although self-reports have the potential to be inaccurate, the CTS uses different time fram es to increase the accuracy o f responses. The women decide whether or not each incident has occurred in the past 6 months, 12 months, or 5 years. This study is interested in the five year category, as it represents the longest period o f tim e during w hich the children were exposed to wife abuse. Once again, a precise estimate o f the number o f violent incidents is not necessary, because the scales w ill be analyzed regarding the total number o f verbally aggressive, mildly violent, and severely violent incidents as opposed to the specific num ber o f each incident. Inform ation about the context in which violent incidents occurred is not required for this study. Children exposed to wife abuse are unable to establish or understand its context, and I do not require this information to support or refute my hypotheses regarding exposure to w ife abuse and the presence o f traditional gender beliefs. Compared to alternative measures o f family violence, the CTS displays higher reliability and validity, and greater scope. Despite the controversy regarding their assessment and interpretation o f violent acts, the CTS continue to dominate research in the family violence field. 43 Adapted Version o f the Conflict Tactics Scales The type, severity, and frequency o f abuse the children were exposed to was assessed by their mothers’ com pletion o f an adapted form o f the Conflict Tactics Scales (CTS). A sim ilar adaptation o f this m easure was used in a study by Jouriles, Barling, and O’Leary (1987) who inserted the phrase, “in front o f (child’s name)” before each CTS description to estim ate the abuse children witnessed. In the present study, the women answered the question, “How many tim es in the past 6 months/12 months/5 years did your partner or ex­ partner do the following in front o f (child’s name) ?”. Reliability^ and validity data are not available on my adaptation o f the CTS. However, sim ilar adaptations have been used in other research w ith positive results, and Straus advocates the use o f adaptations for new research (Straus & Gelles, 1990). In addition, such a m inor adaptation is unlikely to affect the reliability and validity values greatly, especially those as high as the scales o f interest to the present study; the verbal aggression and physical violence scales. Sex Role Inventories Bem Sex Role Inventorv The BSRI has dom inated research in the sex role beliefs field for nearly two decades. It is based on the theory th at men and women judge their own behaviour and personality characteristics according to the differential value society places on traditional male and fem ale traits. 44 Bem (1981) does not consider masculinity and fem ininity to be separate and opposite dimensions. Rather, individuals incorporate varying degrees o f typical male and female characteristics. Depending on the relative strength o f m ale and female traits in an individual’s personality, he or she may be classified as m asculine, feminine, androgynous, or undifferentiated (Bem, 1981). Table 3 outlines the four sex-role dimensions. Table 3: Sex-Role Dim ensions o f the BSRI BSRI Scale Sex-Role Dimension M asculine Fem inine Androgynous Undifferentiated M high low high low F low high high low Respondents who score high on the m asculine scale and low on the feminine scale are considered masculine; those who score high on the fem inine scale and low on the masculine scale are considered fem inine; those who score high on both the masculine and feminine scales are considered androgynous; those who score low on both the masculine and feminine scales are considered imdifferentiated. Psychometric analyses were performed on two sam ples o f undergraduate students at Stanford University (Bem , 1981). I will report the data fiom the more recent 1978 study which included 340 w om en and 476 m en (Bem, 1981). Coefhcient alpha values o f internal consistency were .78 for both m en and women on the feminine scale and .86 for women and .87 for m en on the masculine scale (Bem, 1981). These data support the theoretical proposition that the fem ininity and masculinity scores o f the BSRI are both logically and empirically independent (Bem, 1981). 45 Test-retest reliability^ was assessed a t a four-week follow-up with 28 wom en and 28 men. The analysis demonstrated that test-retest reliability was high for both genders on both the m asculine and feminine scales. The low est value o f .76 occurred for male subjects describing them selves on the m asculine scale ^ e m , 1981). W hen describing them selves on the fem inine scale, men’s test-retest reliability was .89 and women’s was .82 (Bem, 1981). W omen’s self-descriptions on the m asculinity scale showed a value o f .94 (Bem, 1981). The BSRI was also analyzed regarding social desirability, w ith results show ing that subjects did not tend to describe them selves in a socially desirable m anner. Low values o f .03 and .04 for women and men on the fem inine scale and .21 and .02 for women and men on the m asculine scale support the low tendency for subjects to tailor their responses based on social convention or approval (Bem, 1981). Child Sex Role Inventorv The CSRI was developed directly from the widely used and respected BSRI. Both research and personal experience show us that like adults, children hold traditional sex-role beliefs. A t an early age, children recognize that certain characteristics or personality traits are considered more appropriate for one gender than the other. Like adults, children hold these beliefs to varying degrees. This knowledge can affect children’s views o f their own genderrelated personality characteristics and behaviours and influence their relationships w ith others (Boldizar, 1991). Reliability o f the masculine and fem inine scales o f the CSRI was assessed through alpha coefflcients o f internal consistency. Coefficient alpha for the masculine scale was .75, 46 and for the feminine scale was .84 (Boldizar, 1991). Stable test-retest reliabilities o f .71 for the feminine scale and .56 for the m asculine scale were found after a one-year follow-up (Boldizar, 1991). Validity o f the scales was evident in significant gender differences on both scales and in relationships between gender-role categories and measures o f a) sex-typed toy and activity preferences; b) self-perceptions o f global self-w orth, scholastic competence, social acceptance, athletic competence, physical attractiveness, and behavioural conduct; and c) cognitive performance (Boldizar, 1991). Due to the recent development o f the CSRI, further study o f its psychom etric properties has not been conducted. However, as each item on the CSRI is directly related to a corresponding BSRI item, their psychometric properties are expected to be comparable. The existing data and the reputation o f the BSRI support the use o f the CSRI in the present study. Sum m ary o f B S R I and CSRI The BSRI and CSRI are composed o f 20 masculine items, 20 fem inine item s, and 20 neutral items. Each item is a statement about the self, and respondents rate each item according to “how true o f you” it is on a four-point scale. Table 4 lists sample BSRI items and the corresponding CSRI items. M asculinity and femininity scores are calculated by averaging the responses to the 20 items on each scale, thus providing scores ranging from 4 (highest) to 1 (lowest). 47 Table 4: Sample Items from the BSRI and CSRI BSRI Item CSRI Item Sex-Role Dimension Com petitive Compassionate Sincere Analytical Loves children Friendly When I play games, I really like to win. I care about what happens to others. I am an honest person. I like to think about and solve problems. I like babies and sm all children a lot. I have many Mends. Masculine Feminine Neutral Masculine Feminine Neutral In this study, the BSRI and CSRI determined the relative flexibility and rigidity o f the m others’ and children’s sex-role beliefs. Specifically, they determined whether the children exposed to wife abuse were more likely to describe themselves as traditionally feminine than those who had not been exposed to wife abuse. Setting All families in the treatm ent group and one family in the comparison group com pleted their questionnaires at Agency A w ith the child and m other in separate rooms. The other six fam ilies in the comparison group completed their questionnaires w ith the child in a sem inar room at the participating school and the woman at a desk in the hallway. One fam ily in the com parison group completed their questionnaires at the University o f N orthern B ritish Columbia. The rating scale o f the CSRI was posted on the wall o f the interview room at all locations. 48 Procedure The letters distributed to potential participants in the treatm ent group explained that the study’s intent was to assess gender beliefs o f children exposed to wife abuse (Appendix B). W omen interested in participating contacted me by phone, and were inform ed that they would receive 20 dollars to cover transportation and child care costs incurred by participation. They were invited to read the results o f the study upon its com pletion. During this conversation, a tim e was arranged to discuss and sign the informed consent form s and complete the questionnaires. Upon arrival at both Agency A and the elem entary school, I greeted the fam ily in the reception area. We then went to a room where I described the study to the fam ily. Each woman had been informed in the letter and by telephone that the study involved w ife abuse, but the children were unaware that their participation in the study involved exposure to abuse. I told the children that the study investigated how children describe their personality characteristics. All family members signed informed consent forms once they had read, discussed, and understood them (see Appendix E). I then escorted the woman to another room and provided verbal instructions for completing the CTS (Straus, 1979), the adapted version o f the CTS, and the BSRI (Bem, 1981). I returned to the child, and followed a general script for completion o f the CSRI (see Appendix D). I read the CSRI item s and marked the child’s answers on the form to elim inate possible misunderstandings or confusion caused by varying reading abilities o f the children. 49 I f m ore than one child was participating in the study, the other(s) viewed a children’s video in the reception area or played outside while their mother and sibling completed their questionnaires. Each child’s interview took ^proxim ately 20 minutes. The tim e required for the women to complete their questionnaires ranged from IS to 40 minutes, depending on the number o f her children participating. A fter completing their questionnaires, each child was asked how he or she felt about the study, and if he or she had any questions or comments. If so, they were discussed at this tim e. The child then returned to the reception area or played outside while I spoke to his or her mother. The woman and I discussed any negative feelings, questions, or comments that arose while completing the questionnaires. After I discussed the experience with all family members, the children in the treatm ent group were given a pen and their mothers received the payment. Children in the comparison group each received a two dollar McDonald’s gift certificate and a mini-stamper o f their choice. All the women were informed that the results o f the study would be available at Agency A. Collection, Recording, and Analysis o f Data I compiled the questionnaires using M icrosoft Word for Windows Version 6.0 (1994). The women responded to each item on the questionnaires by marking their answers in the appropriate square and I marked the children’s on the paper for them. After coding and grouping the data, I recorded it in tabular form using Microsoft Excel for Windows Version 50 5.0 (1994). The same software was used for all statistical analysis including calculation o f descriptive statistics and m ultiple independent samples t-tests. Methodological Assumptions Like in any research, several assumptions were made in conducting my study. Firstly, I assumed that the treatm ent group o f children exposed to wife abuse I obtained, though not random ly selected, shared characteristics with the population o f children exposed to wife abuse which enabled me to make some preliminary analyses o f the relationship between exposure to wife abuse and the development o f traditional sex-role beliefs. Secondly, I assumed that the information provided by the women and children in my study was accurate and honest, although I acknowledge that retrospective, second-hand accounts o f children’s exposure to wife abuse will be inherently flawed to some degree. Thirdly, I assumed that the instruments I chose would be reliable and valid tools in the assessment o f demographic characteristics, experiences o f wife abuse, children’s exposure to wife abuse, and participants’ sex-role beliefs. Limitations Because this is an ex post-facto study using a purposive sample, it will not establish causality. Children are exposed to traditional sex-role models in every aspect o f their lives. Television programs, movies, music, and peers all contribute to and affect their beliefs about gender, and it is not possible to say whether exposure to wife abuse was the main influence on the development o f their sex-role beliefs. However, the inclusion o f a comparison group 51 allowed me to compare children living in sim ilar social circumstances, whose only m ajor difference was exposure to wife abuse. Due to the sm all, non-representative sample, the results o f this study may be m ost applicable to a lim ited segm ent o f the population o f children exposed to w ife abuse - those between the ages o f 6 and 12 years, who live with their biological mothers, and who are no longer e^gosed to wife abuse in their homes. The fam ilies m et certain demographic criteria, and were predom inantly o f Caucasian and Aboriginal heritage. W ith the exception o f a few participants, m ost o f the participants were from low to middle incom e families, and many o f the women had experienced divorce or separation. The participants in this study were all volunteers who participated based on a belief in hirthering our understanding o f the difhculties children exposed to wife abuse experience or who felt that their fam ily w ould somehow benefit fi-om participation. These people m ay be different in some way firom those who chose not to participate. M ost o f the treatm ent group participants were identified through involvem ent in mental health services. This factor may further decrease generalizability, as these fam ilies m ay be m ore aware o f the effects o f wife abuse on fam ily members, or m ay be at a different stage regarding their acknowledgment o f and openness about wife abuse in their lives. W hile it w ould have increased the accuracy o f estim ates o f children’s exposure to wife abuse and provided inform ation about the nature o f the abusive relationship, inform ation was not obtained from the violent partners or ex-partners in this study, as I believed it would compromise the safety o f the participants. 52 Further study o f this group o f children has not been planned, thus the present study will not provide a longitudinal analysis o f the sex-role beliefs o f these children. They m ay alter their sex-role beliefs as they enter adolescence, becoming either more flexible or more rigid. This study w as a prelim inary step in the sex-role belief research o f a very im portant and oflen neglected group o f children - children exposed to wife abuse. This group o f children is being identified in increasing numbers, and w hile it is not clear whether the incidence o f wife abuse and children’s exposure to it is increasing or whether it is being reported in greater num bers, these children deserve immediate attention and understan d in g . Establishing a relationship between children’s exposure to wife abuse and the development o f traditional sex-role beliefs w ill enable society to help these children develop more positive and egalitarian sex-role beliefs, thereby decreasing the risk o f their being either victims or perpetrators o f violence in their fiiture relationships. 53 CHAPTER FOUR: RESULTS Participants I collected data from a group o f 15 mothers and their 24 children. Twelve children who had been exposed to the abuse o f their m others w ithin the past five years com prised the treatm ent portion o f the sample. The comparison group consisted o f 12 children who had not been exposed to the abuse o f their mothers at any point in their lives. Children between the ages o f 6 and 12 years participated in the study, as research has shown that children in this age range have knowledge o f gender stereotypes (Serbin, Powlishta, & Gulko, 1993). Before the age o f six, children’s gender knowledge is lim ited to categorization o f people as male or female, preferences for sex-typed toys, and a preference for sam e-sex peers (Serbin et al., 1993, p. 15). A study o f 558 children between the ages o f 5 and 12 found increases in knowledge o f stereotypes, flexibility o f stereotypes, and sex-typed personal preferences as the children aged (Serbin et al., 1993, p. v). Adolescents’ gender knowledge reflects increasing flexibility and awareness that sex roles are not rigid or absolute (Serbin et al., 1993, p. 11). Using children between 6 and 12 years ensured relatively consistent gender knowledge among subjects which was not significantly complicated by cognitive developmental differences or pubertal influences. Treatment Group Participants The treatm ent group included children who were exposed to the verbal, emotional, or physical abuse o f their mothers. The children included eight boys and four girls ranging in age from 7 to 12 years, w ith a mean % e o f 9.4 years (SD = 1.5, mdn = 9.7). The mean age o f 54 the boys was 9.2 years (SD = 1.0. mdn = 9.8), while the mean age o f the girls was 8.9 years (SD = 1.6, mdn = 8.9). Comparison Group Participants The comparison group o f 12 six to twelve year-old children participated on the basis o f not living in families characterized by wife abuse. The eight boys and four girls ranged in age from 6 to 12 years w ith a m ean o f 8.9 years (SD = 2.2, mdn = 9.6). The mean age o f the boys was 9.0 years (SD = 2.1, mdn = 8.6), while the mean age o f the girls was 11.5 years (SD = 1.5, mdn = 12.0). Sim ilarity o f Treatment and Com parison Group Children The children in the treatm ent and comparison groups were sim ilar in age and grade level, although the mean age o f the girls in the comparison group was 2.4 years higher than the m ean age o f girls in the treatm ent group. M others o f Children in the Treatm ent Group Table 5 outlines the characteristics o f the mothers o f the children in the treatm ent and comparison groups. The m others o f the children in the treatm ent group were between the ages o f 25 and 44 years (M = 36, SD = 5.6, mdn = 37). One o f the women was single, two were separated, two were divorced, and two were married. The women’s aimual family incomes ranged from $9,000 to 75,500 (M = $29,357, SD = $24,632, mdn = $15,500). Two o f the women were receiving income assistance. O f the seven women, three had received high school diplomas, three continued on to trade school, college, or university, and one 55 reached the post-graduate level. The num ber o f abusive relationships the women experienced ranged from one to six (M = 2.1, SD = 1.8, mdn —2) and the num ber o f years spent in abusive relationships ranged from 2 to 12 (M = 9.6, SD = 4.2, mdn = 12). Two o f the women had been exposed to the abuse o f their own m others as children. Three reported that their partners had been exposed to the abuse o f their mothers; four did not know whether their partners had been exposed to wife abuse as children. A ll o f the wom en indicated that they were not in an abusive relationship at the tim e o f the interview, and none o f the abused women was currently involved w ith a past abusive partner. M others o f Children in the Comparison Group The mothers o f the children in the com parison group were betw een the ages o f 30 and 49 (M = 38, SD = 6.4, mdn = 37). Four o f the women were separated, three were married, and one was widowed. All o f the women were currently involved in a relationship, and none w as receiving income assistance. Their annual fam ily incomes ranged from $45,500 to 75,500 (M = $58,000, SD = $10,350, mdn = $55,500). O f the eight women, three had received high school diplomas, and five continued on to trade school, college, or universi^. None o f the women initially classified her relationship as abusive, but afier com pleting the questionnaires, one woman realized that her daughters had been exposed to verbal abuse against her by their father (see Table 3). A fier considering possible complications and ramifications o f excluding this participant, I decided to keep the woman in the com parison group because although she reported having experienced verbal abuse, she did not report any incidents o f her children’s exposure to incidents o f m ild or severe physical 56 violence. This fact separated her children firom the treatm ent group participants, who had all been exposed to at least one incident o f the physical abuse o f their mothers. Three o f the women had been exposed to the abuse o f their own mothers as children. Six reported that their partners had not been exposed to the abuse of their m others; two did not know whether their partners had been exposed to wife abuse as children. Table 5: Characteristics o f the Mothers o f Children in the Treatment and Comparison Groups Characteristic Treatment Group n=7 Comparison Group n=8 Age A /= 36 A /= 38 M arital Status 1 single 2 separated 2 divorced 2 married 0 widowed 0 single 4 separated 0 divorced 3 married 1 widowed A n n u a ] Family Income M = $29,357 wd>i = $15,500 M = $58,000 mdn = $55,500 Num ber receiving social assistance 2 0 Education Level 3 high school 3 post secondary 1 post-graduate 3 high school 5 post-secondary 0 post-graduate Num ber o f abusive relationships M =2 M= 0 Exposed to wife ahuse as a child 2 3 Partner exposed to wife abuse as a child 3 0 57 Sim ilarity o f M others in the Treatm ent and Comparison Groups As Table 5 shows, the m others o f the children in the treatm ent and com parison groups appear to be sim ilar regarding age, m arital status, and education level. Differences appear in a n n u a l fam ily income and use o f social assistance. The large disparity betw een the median incomes o f the two groups may be due to the presence o f a second income earner in all eight o f the comparison group families, compared to only four o f the treatm ent group families. The women also differed w ith respect to their partners’ exposure to wife abuse as a child. W hile three o f the women who had experienced wife abuse reported that their partners had been exposed to wife abuse as children, none o f the women who had not experienced wife abuse reported their partners’ exposure to wife abuse as children. Estimates o f Children’s Exposure to Wife Abuse The children’s mothers indicated that their children had been exposed to at least one incident o f verbal, emotional, or physical abuse against them w ithin the past five years. Table 6 lists the m others’ estim ates o f each child’s exposure to incidents o f w ife abuse. Median scores o f each item on the adapted version o f the CTS were added together to provide an estim ate o f the total number o f incidents o f verbal aggression, m ild violence, and severe violence to which each child was exposed. The num ber of incidents o f wife abuse each child in the treatm ent group was exposed to ranged from 10 to 270 (A /= 142, SD = 91.4, mdn = 140). O f the three types o f abuse listed in the CTS (verbal abuse, m inor violence, and severe violence), the children had been exposed to verbal abuse m ost fi-equently, including insults and threats o f physical violence against their mothers. The children had been exposed 58 to between 8 and 130 incidents o f verbal aggression (M = 91.3, SD = 46.1, mdn = 112). However, m any o f the children had been exposed to both m ild and severe physical violence, ranging in severity from objects being thrown at their m others to their mothers’ being choked, beaten up, or assaulted w ith a knife or gun by a m ale partner. Incidents o f m ild violence children were exposed to ranged 6om 1 to 75 (M = 32, SD = 29.3, mdn = 25), w hile incidents o f severe violence ranged from 0 to 83 (M = 18.3, SD = 31.0, mdn = 0.5). The treatm ent group children appear to form two groups regarding their exposure to severe violence: children who have been exposed to repeated incidents o f severe violence, and children who have not been exposed to any severe violence. Table 6: Treatm ent and Comparison Group Children’s Exposure to W ife Abuse Verbal Aggression T C (n = 1 2 ) (n = 12) 84 130 24 24 112 112 130 112 100 8 130 130 0 4 0 0 0 0 0 0 0 0 73* 73* M ild Violence T C (n = 12) (n = 12) 8 25 1 1 75 75 55 8 25 1 55 55 0 0 0 0 0 0 0 0 0 0 0 0 Severe Violence T C (n = 12) (n = 12) 0 0 0 0 82 83 18 0 0 1 18 18 0 0 0 0 0 0 0 0 0 0 0 0 Total Violence T C (n = 1 2 ) (n = 12) 92 155 25 25 269 270 203 120 125 10 203 203 * incidents o f verbal aggression against one woman in the com parison group 0 4 0 0 0 0 0 0 0 0 73* 73* 59 BSRI and CSRI Classification o f M asculinity and Femininity Because the BSRI has been standardised, I was able to use median scores o f the norm ative sam ple to place the mothers o f the child participants in one o f the four BSRI sexrole categories previously summarized in Table 3. I used Bem’s standardized t scores to classify the children’s m others as masculine, fem inine, androgynous, or im differentiated on the BSRI (1981). W omen scoring high on the masculine scale and low on the fem inine scale were classified as m asculine, those scoring high on the feminine scale and low on the m asculine scale were classified as fem inine, those scoring high on both the m asculine and fem inine scales were classified as androgynous, and those scoring low on both the m asculine and fem inine scales were classified as undifferentiated. The CSRI is a relatively new instrum ent for which norm ative data are not available. As it has not been standardized, I focused on an analysis o f the children’s self-reported m asculinity or fem ininity and did not classify child subjects as androgynous or undifferentiated. Table 7 outlines the score distribution used to classify children as either m asculine or fem inine. Table 7: Sex-Role Categories o f the CSRI CSRI Scale Sex-Role Classification M asculine Fem inine M asculine high low Feminine low high Children who scored high on the masculine scale and low on the fem inine scale were considered m asculine because they endorsed those qualities typically considered traditionally 60 m asculine. Those who scored high on the fem inine scale and low on the masculine scale were considered feminine because they endorsed those qualities typically considered traditionally feminine. For the purposes o f this study, it was sufficient to classify the children as either m asculine or feminine in order to determine whether children exposed to wife abuse hold m ore traditional sex-role beliefs than children not exposed to wife abuse. Data Analysis Given my small sample size and the exploratory nature o f my study, I chose to analyze m y data using multiple t-tests with a significance level o f .10. I commented when results were significant at the more rigorous .05 level. I calculated Cohen’s effect size (d) for each o f the t-test analyses in order to clarify the significance o f differences between the m eans o f the sm all samples (Glass & Hopkins, 1984). These values more clearly illustrate the m agnitude o f any differences found between groups and are presented in standard deviation units in Tables 8 through 17. Women’s BSRI Scores I initially ran t-tests on the women’s BSRI data to determine the degree to w hich they described themselves as masculine or feminine. The results o f these analyses are presented in Tables 8 and 9. W omen who had experienced wife abuse held less traditional beliefs regarding m asculinity and femininity than women who had not experienced wife abuse. I then used the median-split method to classify each o f the women as m asculine, fem inine, androgynous, or undifferentiated, allowing me to develop a greater understanding o f the women’s self-described masculinity and femininity. 61 W omen’s BSRI MaamHnitv Scores I analyzed the BSRI scores o f the m others o f the treatm ent group and com parison group children to determine whether women who had experienced wife abuse described them selves as less masculine than women who had not experienced wife abuse. Table 8 presents the results o f the analysis, illustrating that women who had experienced wife abuse were signiGcantly different from women who had not experienced w ife abuse in the extent to which they described themselves as masculine (/(13) = —2.41, p < .03). The m asculinity scores o f women who had experienced wife abuse were 1.25 SD below those o f women who had not experienced wife abuse. Women who had experienced wife abuse described themselves as significantly less masculine than those who had not experienced wife abuse. Table 8: BSRI M asculinitv Scores o f Treatm ent and Comparison Group M others Treatm ent n=7 73.29* * scores out o f 140 * p < .10 ** p < .05 Comparison n —8 88.63 Effect Size /(df) = 13 -2 .4 1 * * (d) -1 .2 5 W omen’s BSRI Fem ininity Scores I analyzed the BSRI scores o f the m others o f the treatment group and com parison group children to determine whether women who had experienced wife abuse described them selves as more feminine than women who had not experienced wife abuse. Table 9 presents the results o f the analysis, illustrating that women who had experienced wife abuse were significantly different from women who had not experienced 62 wife abuse in the extent to which they described them selves as feminine (/(13) = —1.89, p < .08). The fem ininity scores o f women who had experienced wife abuse were 0.98 SD below those o f wom en who had not experienced wife abuse. W omen who had experienced wife abuse rated them selves as significantly less fem inine than women who had not experienced wife abuse. Table 9: BSRI Fem ininitv Scores o f Treatm ent vs. Com parison Group Mothers Treatm ent n=7 95.71' scores out o f 140 Comparison n=8 105.25 Effect Size t(dQ = 13 -1 .8 9 * (4 -0 .9 8 ' * p < .\ 0 ** p < .0 5 Sum m ary o f W omen’s BSRI Results The statistical analyses described above outline the differences in sex-role beliefs between wom en who have experienced wife abuse and those who have not. Table 10 summarizes the analyses, illustrating the pattern o f significant findings on both the masculine and fem inine scales o f the BSRI. Table 10: S um m ary o f Treatment and Comparison Group M others’ BSRI Results Group Treatm ent Group Com parison Group Sig. Level M F p < .0 5 p < .1 0 Five o f the women who had experienced wife abuse scored low on both the masculine and fem inine scales, and are therefore considered undifferentiated on the BSRI. The other 63 two wom en who had experienced wife abuse scored higher on the feminine scale than the m asculine scale and are therefore considered feminine. O f the women who had not experienced wife abuse, three scored low on both the m asculine and feminine scales, and are considered undifferentiated on the BSRI. Two wom en scored high on both the masculine and feminine scales, and are considered androgynous; three women scored high on the feminine scale and low on the m asculine scale, and are considered feminine on the BSRI. Children’s CSRI Scores The results o f my study o f children’s sex-role beliefs are summarized in Tables 11 through 17. The CSRI differentiated between children exposed to wife abuse (EW A) and children not exposed to wife abuse (NEWA) in three o f the statistical analyses. The treatm ent groups o f children exposed to wife abuse and boys exposed to wife abuse described them selves as less feminine than the comparison group. No significant differences were discovered regarding children’s self-described masculinity in either the treatm ent or com parison groups or in either gender. H vpothesis 1: Bovs Exposed to W ife Abuse vs. Bovs Not Exposed to W ife Abuse I analyzed the CSRI scores o f boys exposed to wife abuse and boys not exposed to wife abuse to determine whether boys exposed to wife abuse described them selves as more m asculine and less feminine than boys not exposed to wife abuse. Table 11 summarizes the results o f the analysis o f boys’ masculinity scores. Boys exposed to wife abuse were not significantly different fiom boys not exposed to w ife abuse in 64 the extent to which they described themselves as masculine (t(20) = —l.56,p < .43). The m asculinity scores o f boys exposed to wife abuse were 0.78 SD below those o f boys not exposed to wife abuse. Table 11 : CSRI Masculinity Scores o f Bovs EWA vs. Bovs NEWA * * EWA n=8 54.25* scores out o f 80 NEWA n=8 60.25 Effect Size /(df) = 14 - 1 .5 6 id) -0 .7 8 parison Treatment Group Figure 3. The apparent disordinal interaction between group and m asculinity score. Boys' and Girls' CSRI Masculinity Scores 62 8 «3 S3 1 GO Î 1 60 58 56 Boys Girls 54 52 50 48 46 Conq)arison Treatm ent Group Figure 4. The apparent ordinal interaction between group and fem ininity score. A lthough significant results were obtained only for femininity scale data, the above figures illustrate sim ilar patterns among sex-role beliefs o f boys and girls in the treatm ent and com parison groups. Boys exposed to wife abuse appear to become less m asculine and less 78 fem inine than boys not exposed to wife abuse, while girls in both groups do not appear to change their sex-role beliefs significantly as a result o f exposure to wife abuse. Bovs Exposed to W ife Abuse vs. Girls Exposed to Wife Abuse Dividing the group o f children exposed to wife abuse into boys and girls led once again to significant findings on the femininity scale. Compared to girls exposed to wife abuse, boys exposed to wife abuse described themselves as less feminine (See Figure 4). In fam ilies characterized by wife abuse, it is common to see the male abuser belittle or criticize his w ife as he sim ultaneously attempts to present a masculine image o f power, domination, and control. Children exposed to wife abuse, in their lower self-reported fem ininity scores, may be reflecting this pattern o f devaluing traditionally feminine characteristics. It is not necessary that they display a parallel increase in masculinity, because the sex-role inventory does not use a bipolar scale for assessing m asculinity and femininity. High scores on one sex-role dimension does not necessarily result in low scores on the other. Integration o f Findings w ith Past Literature The intergenerational transmission o f violence theory has been studied by numerous fam ily violence researchers, often w ith inconclusive results. N ot surprisingly, this prelim inary study failed to answer the question o f whether abuse is transm itted across generations. However, it does point to the need for continued research adding to the existing literature which suggests that boys and girls exposed to wife abuse may incorporate their parents’ sex role beliefs, placing them at risk o f being involved in abusive relationships as 79 adults (Brassard et al., 1991; Health and W elfare Canada, 1992; Kashani et al., 1992; Moore et al., 1990; Osofsky, 1995: T ^ p and Hinish, 1992). Contribution o f Findings to Literature M y findings, though lim ited to significant results on the feminine scale, clearly support continued research on the sex-role beliefs o f children exposed to wife abuse. I have illustrated that children exposed to wife abuse are by no means immune to the repeated scenes o f violence they are exposed to. Even in a sm all sample, children’s exposure to wife abuse results in the incorporation o f traditional sex-role beliefs among children, especially regarding their views o f femininity. Im plications o f Findings Theoretical Im plications M y study lends fiirther support to contem porary theories o f wife abuse. Social learning theory, gender schem a theory, abusive personality theory, and fem inist theory all suggest that children exposed to wife abuse are at heightened risk o f involvement in relationships characterized by violence. My finding that children’s sex-role beliefs are affected by exposure to wife abuse supports continued use o f these theories as a foundation for further exam ination o f the potential intergenerational transm ission o f violence. Although many theorists argue for the accuracy and strength o f one theory, I believe we m ust incorporate different aspects o f num erous theories o f wife abuse to develop our understanding o f the effects o f children’s exposiue to wife abuse. I do not believe any one 80 theory can adequately address the numerous social, fam ilial, and individual issues involved in the relationship between exposure to wife abuse and traditional sex-role beliefs. R esearch Im plications My study points to the need for continued study o f the population o f children exposed to wife abuse. N ot only are they at risk o f behavioral, physical, and psychological problem s, they m ay incorporate sex-typed beliefs which have the power to negatively affect personal relationships throughout their lives. Longitudinal studies o f large samples o f children exposed to wife abuse would be valuable in this area, as our sex-roles are altered by a num ber o f factors throughout our lives. Age, culture, religious afhliations, fam ily relationships, friendships, dating relationships, m arital partnerships, and life experiences all have the ability to affect the flexibility o f our sex-role beliefs. Research has identified the powerful effects o f wife abuse on its perpetrators and victim s; now we must turn our attention to the children exposed to wife abuse as they m ay sufler the consequences o f exposure to wife abuse throughout their entire lives. They m ay, in the end, inflict the same pain on their own partners and children. Practical Im plications As young people, children are an excellent target audience for educational and preventative efforts regarding abuse in intim ate relationships. Children are willing and able to change their beliefs and adopt new, healthier sex-role beliefs. Practitioners can thus focus on children as the generation to develop healthier, more balanced, egalitarian sex-role beliefs. 81 Because modelling is such a powerful influence in children’s lives, practitioners can use it to decrease traditional sex-role beliefs regarding power and control, and enhance those o f equality and respect. Counselling groups for children exposed to wife abuse are an etifective method o f providing such modelling; fellow group members and the group facilitators provide examples o f different ways o f thinking, feeling, and acting. Groups are also a valuable forum for modelling conflict resolution, anger management, and communication skills - areas which children exposed to wife abuse often lack skills in. Although the effects o f exposure to wife abuse on children’s sex-role beliefs was only partially supported in this study, counsellors and parents need to be aware o f the potential damaging effects o f children’s incorporation o f traditional sex-role beliefs. Counsellors can support fam ilies characterized by wife abuse develop increased understanding o f the effects o f exposure to wife abuse on children. Parents could then recognize these effects in their children, and better understand the power and influence o f their actions, attitudes, and beliefs on their children. W hen working w ith a family characterized by wife abuse, it is valuable to focus on the effects o f society’s devaluing o f women and traditional feminine characteristics. Society as a w hole has historically placed greater status and privilege on men, and many practitioners in the m ental health and educational fields have the opportunity to address such differential treatm ent based on gender. 82 Limitations M ethodological T.imitations The most obvious lim itation o f my study was also the most unavoidable - the small sample size. Despite extensive efforts at recruiting a larger sample, I was unable to do so. Several factors contributed to this problem , including the controversial and em otional nature o f my topic, difSculty accessing women who have experienced wife abuse w ithout encroaching on their rights to privacy, the financial and social stressors these wom en often experience which make it difficult to participate in such a project, and m y own financial lim itations and time constraints. Although the small sample size lim ited my ability to identify significant differences among my treatment and comparison groups, I succeeded in identifying some important relationships among gender, exposure to w ife abuse, and sex-role beliefs. The 15 families who participated in my study did so voluntarily. As w ith any research study, volunteers may be different than those who choose not to participate. Those who volunteered for the treatment group m ay have been more open about their experiences o f wife abuse, or may have spent more tim e out o f the abusive relationship than those who chose not to participate. Women in the comparison group had very little to gain fi*om participating, and may therefore have been more informed or concerned about the issue o f wife abuse than those who chose not to participate. M ost o f the women agreed to their fam ilies’ participation before being informed o f the 20 dollar imbursement; only one woman was informed o f the money before agreeing to 83 participate. It is possible that the m oney provided some incentive for her to participate. However, I did not feel that the 20 dollar sum was large enough to be coercive, and 1 feel that even though the treatm ent group m others expressed appreciation for the m oney, all would have participated w ithout any monetary benefit. W hen using a clinical sample, there is always the possibility that the treatm ent group is different firom the population being studied. The treatment group fam ilies had sought m ental health services at one or more community agencies, and may therefore have been at a différent stage regarding their acknowledgment and understanding o f their experiences o f wife abuse than those who chose not to participate. They have also have had m ore social, financial, or emotional resources than the families who had experienced w ife abuse and chose not to participate. An additional lim itation which 1 did not assess is the extent to w hich the treatm ent group women’s and children’s sex-role beliefs may have been altered by their involvem ent in counselling or support groups. 1 do not know whether their individual or group sessions addressed sex-roles or gender beliefs, but as far as 1 know, these issues w ere not specifically incorporated in their counselling programs. Generalizabilitv Any small, ex post-facto study is lim ited in the extent to which it can form generalizations about the population it studies. My study was based on very specific treatm ent and comparison groups, ones which consisted o f children betw een the ages o f 6 and 12, who were firom low to middle-income families. M ost o f the treatm ent group fam ilies 84 were sole parent fam ilies headed by a woman. None o f the children was currently living w ith their m others’ past abusive partner, who, in m ost cases, was the child’s father. My sam ple was selected firom a medium sized, resource-based community in northern British Columbia. The results o f my study can therefore not be generalized to the population o f children exposed to wife abuse. However, I feel that regardless o f their age, socioeconomic status, single or dual parent status, and place o f residence, fam ilies characterized by wife abuse share m any common characteristics. Home environm ents o f fear, uncertainty, and anxiety are typical, and the population o f children exposed to wife abuse share m any o f their early experiences o f intim ate relationships. Future Directions An ideal study o f sex-role beliefs and children’s exposure to wife abuse would include a large sample o f boys and girls o f various ages whose sex-role beliefs could be assessed at different stages in their development. Longitudinal studies could clarify those changes in sex-role beliefs which are affected by puberty, cognitive development, peer relationships, the media, and life experiences such as exposure to wife abuse. Studies o f this type can also address social, religious, and cultural factors which can not always be addressed in sm aller, short-term studies. Although much fam ily violence research focuses on the obvious negative effects o f exposure to wife abuse on boys, I advocate a greater focus on the effects on girls, as they are equally at risk o f lifelong, negative consequences o f exposure to wife abuse. N ot only are women the m ost firequent victims o f fam ily violence, they are also undervalued and 85 discrim inated in society as a whole. This fact is reflected in the research on wife abuse itself, as girls’ responses to exposure to wife abuse are minimized, apparently because they are not as “externalized” or obvious as those o f boys (Jaffe, Wolfe, & W ilson, 1990). I feel that whether children react to traum a aggressively or passively does not alter the significance o f their experiences, and focusing research efforts on one gender at the expense o f the other is a m istake which could affect generations to come. Continued research in this area would benefit fiom gathering information fiom as many fam ily members as possible. M y study relied on the women’s estimates o f their children’s exposure to wife abuse, but gathering information fiom the abusive partners would be extremely valuable. When feasible, children exposed to wife abuse could also provide estim ates o f their exposure to wife abuse, as mothers and fathers are often unaware o f many incidents witnessed by their children. Specific information on the m en’s own sex-role beliefs would add considerable depth to any study, as it is beliefs about power, dominance, and control which ofien lead to abuse. As the BSRI was developed nearly 20 years ago, it is appropriate and advisable to seek or develop instruments which are more contemporary in their descriptions o f m asculinity and femininity. I chose the BSRI because it provided the foundation for the recent CSRI and is widely respected for its reliability and validity, but I feel that as awareness o f this issue increases, instruments with a more current outlook w ill deepen our understanding. Qualitative studies o f this population would be the m ost direct and thorough way o f assessing children’s sex-role beliefs and experiences o f exposure to wife abuse. Adults often 86 underestimate the extent o f children’s understanding, and I feel that children exposed to wife abuse are a resource w hich is being neglected to a large degree. There are obviously many hurdles to overcome in working directly w ith these children, but overcoming them would be invaluable. Continued research in the area o f sex-role beliefs and children exposed to wife abuse has lim itless potential for increasing our understanding o f this group o f children and enabling practitioners and parents to develop effective ways o f helping them identify or prevent the variety o f behavioural, physical, psychological effects they often experience. We can also leam to identify and address potentially damaging sex-role beliefs in an effort to minimize the likelihood o f their becom ing involved in abusive relationships as adults. 87 Footnotes ' As I was unable to obtain a commercial copy o f the C hild Sex Role Inventory, I com piled this questionnaire using Boldizar’s article, “A ssessing sex typing and androgyny in children: The children’s sex role inventory” (Boldizar, 1991). The article contains the 60 questions com prising the CSRI, as well as instructions for its distribution and scoring. 88 References Bagarozzi, D. A., & Schumm, W. R, (1989). Family measurement techniques. The American Journal o f Family Therapy, 17(2), 165-168. Barnett, O. W., M iller-Perrin, C. L., & Perrin, R. D. (1997). Family Violence Across the Lifespan. London: Sage. Basow, S. A. (1992). Gender Stereotypes and Roles (3rd ed.). Pacific Grove, CA: Brooks/Cole. Becvar, D. S., & Becvar, R. J. (1996). Family Therapy: A Systemic Integration (3rd ed.). Boston: A llyn & Bacon. Bem, S. L. (1981). Bem Sex Role Inventory Perm ission Set. Palo Alto, CA: Consulting Psychologists Press. Boldizar, J. P. (1991). Assessing sex typing and androgyny in children: The children’s sex role inventory. Developmental Psychology, 27(3), 505-515. Brassard, M. R., Hart, S. N ., & Hardy, D. B. (1991). Psychological and em otional abuse o f children. In R. T. Ammerman & M. Hersen (Eds.), Case Studies in Family Violence (pp. 255-270). New York: Plenum Press. Cantos, A. L., Neidig, P. H., & O’Leary, D. (1994). Injuries o f women and m en in a treatm ent program for dom estic violence. Journal o f Family Violence, 9(2), 113-124. Choice, P., Lamke, L. K., & Pittman, J. F. (1995). Conflict resolution strategies and m arital distress as m ediating factors in the link between witnessing interparentàl violence and wife battering. Violence and Victims, 70(2), 107-119. Copping, V. E. (1996). Beyond over- and under-control: Behavioural observations o f shelter children. Journal o f Family Violence, 11(1), 41-57. DeKeserdy, W. S., & Kelly, K. (1993). Woman abuse in university and college dating relationships: The contribution o f the ideology o f fam ilial patriarchy. Journal o f Human Justice, 4(2), 25-52. Dobash, R. E., & Dobash, R. P. (1992). Women, Violence, and Social Change. New York: Routledge. 89 Dobash, R. P., Dobash, R. E., Wilson, M ., & Daly, M. (1992). The m yth o f sexual symmetry in m arital violence. Social Problems, 39(1), 71-91. Doumas, D., Margolin, G., & John, R. S. (1994). The intergenerational transm ission o f aggression across three generations. Journal o f Family Violence, 9(2), 157-175. Dutton, D. G. (1995). The Domestic Assault o f Women. Vancouver, Canada: UBC Press. Dutton, D. G., Starzomski, A ., & Ryan, L. (1996). Antecedents o f abusive personality and abusive behaviour in wife assaulters. Journal o f Family Violence, 11(2), 113-132. Fantuzzo, J. W ., DePaola, L. M ., Lambert, L., M artino, T., Anderson, G ., & Sutton, S. (1991). Effects o f interparental violence on the psychological adjustment and competencies o f young children. Journal of Consulting and Clinical Psychology, 59{2), 258-265. G elles, R. J., & Cornell, C. P. (1990). Intimate Violence in Families (2nd ed.). Newbury Park, CA: Sage. G elles, R. J. (1993). Through a sociological lens: social structure and fam ily violence. In R. J. Gelles & D. R. Loseke (Eds.), Current Controversies on Family Violence (pp.31-46). Newbury Park, CA: Sage. G lass, G., & Hopkins, K. D. (1984). Statistical Methods in Education and Psychology (2nd Ed.). Boston: A llyn & Bacon. Groves, B. M ., Zuckerman, B., Marans, S., & Cohen, D. J. (1993). Silent victims: Children who witness violence. Journal o f the American Medical Association, 269(2), 262264. Grych, J. H., Seid, M., & Fincham, F. D. (1992). Assessing m arital conflict from the child’s perspective: The children’s perception o f interparental conflict scale. Child Development, 62, 558-572. H ealth and W elfare Canada. (1992). Wife abuse - The impact on children. Ontario, Canada: N ational Clearinghouse on Family Violence. Henning, K., Leitenberg, H., Covery, P., Turner, T., & Bennett, R.T. (1996). Long­ term psychological and social impact o f witnessing physical conflict between parents. Journal o f Interpersonal Violence, 11(\), 35-51. 90 Hotaling, G. T., & Sugannan, D. B. (1990). A risk m arker analysis o f assaulted wives. Journal o f Family Violence, 5(1), 1-13. Hughes, H. M ., Parkinson, D., & Vargo, M. (1989). W itnessing spouse abuse and experiencing physical abuse: A “double whammy”? Journal o f Family Violence, 4(2), 197209. Hurley, D. J., & JafiTe, P. (1990). Children’s observations o f violence: H. Clinical im plications for children’s m ental health professionals. Canadian Journal o f Psychiatry, 55(6), 471-476. Jacklin, C. N ., & Reynolds, C. (1993). Gender and childhood socialization. In A. E. Beall & R. J. SXcTrA>ct%(^ds.), The Psychology o f Gender (pp. 197-214). New York: Guilford Press. JafiFe, P., Hurley, D. J., & Wolfe, D. (1990). Children’s observations o f violence: I. Critical issues in child development and intervention planning. Canadian Journal o f Psychiatry, 55(6), 466-476. Jaffe, P., W ilson, S. K., & Wolfe, D. (1988). Specific assessm ent and intervention strategies for children exposed to wife battering: Prelim inary empirical investigations. Canadian Journal o f Community Mental Health, 7(2), 157-163. JafiFe, P., W olfe, D., & W ilson, S. K. (1990). Children o f Battered Women. Newbury Park, CA: Sage. Johnson, H. (1996). Dangerous Domains. Toronto: Nelson Canada. Jouriles, E. N ., Barling, J., & O’Leary, K. D. (1987). Predicting child behavior problem s in m aritally violent families. Journal o f Abnormal Child Psychology, 5 7,453-455. Kashani, J. H., Daniel, A. E., Dandoy, A. C., Holcomb, W. R. (1992). Family violence: Impact on children. Journal o f the American Academy o f child and Adolescent Psychiatry, 57(2), 181-189. Kurz, D. (1993). Physical assaults by husbands: A m ajor social problem. In R. J. Gelles & D. R. Loseke (Eds.), Current Controversies on Family Violence (pp. 88-103). Newbury Park, CA: Sage. McCloskey, L. A. (1996). Socioeconomic and coercive power w ithin the family. Gender & Society, 70(4), 449-463. 91 M cCloskey, L. A , Figueredo, A. J., & Koss, M. P. (1995). The e£fects o f systemic fam ily violence on children’s m ental health. Child Development, 66(5), 1239-1261. M cCue, M. L. (1995). Domestic Violence: A Reference Handbook. Santa Barbara, CA: Contemporary W orld Issues. M icrosoft Excel V ersion 5.0 [Computer Software]. (1994). M icrosoft Corporation. M icrosoft W ord V ersion 6.0 [Computer Software]. (1994). M icrosoft Corporation. M oore, T., Pepler, D., W einberg, B ., Hammond, L., Waddell, J., & W eiser, L. (1990). Research on children from violent homes. Canada’s Mental Health, 35(2/3), 19-23. M ullender, A., & M orley, R. (Eds ). (1994). Children Living with Domestic Violence. London, England: W hiting & Birch. O’Brien, M ., John, R. J., M argolin, G., & Erel, O. (1994). Reliability and diagnostic efficacy o f parents’ reports regarding children’s exposure to marital aggression. Violence and Victims, 9(1), 45-62. O ’Keefe, M. (1994). Adjustm ent o f children from maritally violent homes. Families in Society: The Journal o f Contemporary Human Services, 75(7), 403-415. O ’Leary, K. D. (1993). Through a psychological lens: Personality traits, personality disorders, and levels o f violence. In R. J. Gelles & D. R. Loseke (Eds.), Current Controversies on Family Violence (pp. 7-30). Newbury Park, CA: Sage. Osofsky, J. (1995). Children who w itness dom estic violence: The invisible victim s. Social Policy Report: Societyfor Research in Child Development, 9(3), 1-20. Serbin, L. A., Pow lishta, K ., & Gulko, J. (1993). The development o f sex typing in m iddle childhood. Monographs o f the Societyfor Research in Child Development, 58(2, Serial No. 232), v - 95. Smith, M. D. (1990). Patriarchal ideology and wife beating: A test o f the fem inist hypothesis. Violence and Victims, 5(4), 257- 273. Sigler, R. T. (1989). Domestic Violence in Context. Toronto: Lexington Books. Sternberg, K. J., Lamb, M . E., Greenbaum, C., Cicchetti, D., Dawud, S., M anela Cortes, T., Krispin, O., & Lorey, F. (1993). Effects o f domestic violence on children’s behaviour problem s and depression. Developmental Psychology, 29(1), 44-52. 92 Straus, M. (1979). Measuring intrafam ily conflict and violence: The Conflict Tactics Scales. Journal o f Marriage and the Family, 41, 75-88. Straus, M. A., & Gelles, R. J. (1990). Physical Violence in American Families. (C. Smith, Ed.). N ew Brunswick, NJ: Transaction. Suh, E. K., & Mazur-Abel, E. (1990). The impact o f spousal violence on the children o f the abused. Journal o f Independent Social Work, 4(A), 27-34. Tapp, A., & Hinish, P. (1992). Family violence: The impact on children. Results o f a Working Conference on the Community Response to Family Violence. Colorado: Colorado Dept, o f Social Services, Child Welfare Services Division, 20-27. Tomkins, A. J., Mohamed, S., Steinman, M., Macolini, R. M., Kenning, M. K., & Afirank, J. (1994). The plight o f children who wimess woman battering: Psychological knowledge and policy implications. Law and Psychology Review, 1 8 ,137-187. Tutty, L. M., & W agar, J. (1994). The evolution o f a group for young children who have w itnessed fam ily violence. Social Work with Groups, 77(1/2), 89-104. W alker, L. (1979). ITie Battered Woman. New York: Harper & Row. W ilson, M., & Daly, M. (1993). Spousal homicide risk and estrangement. Violence and Victims, 5(1), 3-15. W olfe, D. A., & Jaffe, P. (1991). Child abuse and family violence as determinants o f child psychopathology. Canadian Journal o f Behavioural Science, 23(3), 282-299. Women in Canada: A Statistical Report. (1995). Statistics Canada, Housing, Family, and Social Statistics Division. Ottawa: Statistics Canada. Catalogue #89-503E. Yllo, K. A. (1993). Through a fem inist lens: Gender, power, and violence. In R. J. Gelles & D. R. Loseke (Eds.), Current Controversies on Family Violence (pp. 47-62). Newbury Park, CA: Sage. Zucker, K. J., & Torkos, H. (1987). Assessment o f Androgyny in Children. (Report No. PS 017 293). Baltimore, MD: Society for Research in Child Development. (ERIC Document Reproduction Service No. ED 300 105). 93 Appendix A Instruments Demographic Q uestionnaire This questionnaire is used to gather demographic inform ation about you and your fam ily. Please answ er the following questions as accurately and honestly as possible. 1. Your age: □ 15-19 □ 20-24 □ 25-29 □ 30-34 □ 35-39 □ 40-44 a 45-49 □ 50-54 a 55-59 □ 60-64 □ 65-69 a 70 or older 2. Occupation: _______________________________ □ □ Full tim e Part tim e 3. Annual income: □ □ □ □ <$10,000 $11,000-$20,000 $21,000 - $30,000 $31,000 - $40,000 □ $41,000 - $50,000 □ $51,000 - $60,000 □ $61,000 - $70,000 □ > $70,000 4. Please check the highest level o f education you have completed. □ □ □ □ □ □ elem entary school junior secondary school senior secondary school some university courses imdergraduate university degree some technical training □ technical diplom a □ some graduate courses □ graduate degree □ some doctoral courses □ doctoral degree 94 5. Present m arital status: p lease mark all that apply) □ single □ cohabited □ m arried □ separated □ □ □ □ divorced rem arried widowed o th e r___ □ □ □ 4 5 6 o r more 6. Num ber o f children: □ 1 □ 2 □ 3 7. Num ber o f children under the age o f 19 living in your home: □ 1 □ 2 □ 3 a 4 a 5 □ 6 or more 8. Ages o f your children: (please mark all that apply) N ote: Place the appropriate number o f check marks beside the num ber if you have more than one child o f that age (e.g. twins, stepchildren o f the same age). □ □ □ □ □ □ □ □ □ □ 1 2 3 4 5 6 7 8 9 10 9. Are you currently involved in a relationship? □ Yes □ No a 11 □ □ □ a □ a a a a 12 13 14 15 16 17 18 19 20 + 95 10. If yes, is this relationship emotionally, verbally, physically, or sexually abusive? □ □ Yes No 11. Number o f past and present relationships which have been characterized by verbal, physical, or sexual abuse; □ 0 a4 □ 1 as □ □ □ 6 or more 2 3 emotional, 12. Total num ber o f years spent in emotionally, verbally, physically, or sexually abusive relationships: □ 0 □6 □ □ 7 □8 □ 9 □ 10 □ more than 10 □ □ □ □ 1 2 3 4 5 13. Have you noticed behavioural or emotional changes in your children which seemed to begin after witnessing domestic violence? □ □ Yes No If yes, please describe: 96 Conflict Tactics Scales No m atter how well a couple get along, there are times when they disagree, get annoyed with the other person, or have fights because they’re in a bad mood or tired or to r some other reason. They also use many different ways o f trying to settle their differences. The following list describes some things that you and your partner/ex-partner m ight do when you have an argument. I would like you to tell me how many tim es (Once, Twice, 3-5 times, 610 tim es, 11-20 times, or more than 20 tim es) your partner or ex-partner has done the following things. Please use the categories below when answering the following questions. Never Once Twice 3-5 times 6-10 times 11-20 times 0 1 2 3 4 5 During a conflict with you, how m any tim es has your partner or ex-partner... Discussed an issue calmly Got inform ation to back up his side o f things Brought in, or tried to bring in, someone to help settle things Insulted or swore at you Sulked or refused to talk about an issue Stomped out o f the room, house, or yard Cried Did or said something to spite you Threatened to h it or throw something at you Threw, smashed, hit, or kicked something Threw som ething at you Pushed, grabbed, or shoved you Slapped you Kicked, bit, or h it you with a fist H it or tried to hit you w ith something Beat you up Choked you Threatened you with a knife or gun Used a knife or fired a gun in the past6 months in the past 12 months in the p asts years More than 20 times 6 If you chose 0, has it ever happened ? Y N Y N Y N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y N N N N N N N N N N N N N N N N 97 Adapted Version o f the Conflict Tactics Scales Children often see or hear conflict between their parents. Using the same categories as the previous page, please try to rem em ber how many tim es your partner has done the follow ing things within sight or hearing o f your child. Please complete one o f these pages for each o f your children that is participating in this study. Never Once Twice 3-5 times 6-10 times 11-20 times 0 1 2 3 4 5 During a conflict w ith you, how m any times has your partner or ex-partner done the following in front o f your child (within sight or hearing): Discuss an issue calmly Get information to back up his side o f things Bring in, or try to bring in, someone to help settle things. Insult or swear at you Sulk or refrise to talk about an issue Stomp out o f the room, house, or yard Cry Do or say something to spite you Threaten to hit or throw som ething at you Throw, smash, hit, or kick som ething Throw something at you Push, grab, or shove you Slap you Kick, bite, or hit you w ith a fist H it or try to hit you w ith something Beat you up Choke you Threaten you with a knife or gun Use a knife or fire a gun In the past6 months In the past 12 months In the pasts years More than 20 times 6 If you chose 0, has your child ever seen or heard it? Y N Y N Y N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y N N N N N N N N N N N N N N N N 98 Bem Sex Role Inventory O n the next page, you w ill find a list o f personality characteristics. I would like you to use those characteristics to describe yourself. Please indicate, on a scale from 1 to 7, how true o f you each o f these characteristics is. Please do not leave any characteristic unmarked Exam ple: sly W rite a 1 if it is never or almost never true that you are sly. W rite a 2 if it is usually not true that you are sly. W rite a 3 if it is sometimes but infrequently true that you are sly. W rite a 4 if it is occasionally true that you are sly. W rite a 5 if it is often true that you are sly. W rite a 6 if it is usually true that you are sly. W rite a 7 if it is always or almost alwcQ^s true that you are sly. If you feel it is sometimes but infrequently true that you are “sly”, never or almost never true that you are “m alicious,” always or almost always true that you are “irresponsible,” and often true that you are “carefree,” then you would rate these characteristics as follows: Sly M alicious Irresponsible Carefree 3 1 7 5 99 1 2 3 4 5 6 7 Never or almost never true Usually not true Sometimes but infiequently true Occasionally true Often true Usually true Always or almost always true 1. Defend m y own beliefs 2. Affectionate 3. Conscientious 4. Independent 5. Sympathetic 6. Moody 7. Assertive 8. Sensitive to needs o f others 9. Reliable 10. Strong personality 11. Understanding 12. Jealous 13. Forcefiil 14. Compassionate 15. Truthful 16. Have leadership abilities 17. Eager to soothe hurt feelings 18. Secretive 19. W illing to take risks 20. Warm 21. Adaptable 22. Dominant 23. Tender 24. Conceited 25. W illing to take a stand 26. Love children 27. Tactful 28. Aggressive 29. Gentle 30. Conventional I I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 100 1 2 3 4 5 6 7 Never or almost never true Usually not true Sometimes but infirequently true Occasionally true Often true Usually true Always or almost always true 31. Self-reliant 32. Yielding 33. Helpful 34. Athletic 35. Cheerful 36. Unsystematic 37. Analytical 38. Shy 39. Inefficient 40. M ake decisions easily 41. Flatterable 42. Theatrical 43. Self-sufficient 44. Loyal 45. Happy 46. Individualistic 47. Soft-spoken 48. Unpredictable 49. M asculine 50. Gullible 51. Solemn 52. Competitive 53. Childlike 54. Likable 55. Ambitious 56. Do not use harsh language 57. Sincere 58. Act as a leader 59. Feminine 60. Friendly 1 1 1 1 1 1 1 1 1 1 1 1 1 I 1 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 3 4 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 101 Child Sex Role Inventory This questionnaire asks about how children your age think about them selves. I will read several descriptions to you, and you w ill decide how well each one describes you. Answer “one” if the statem ent is not at all true o f you, “two” if the statem ent is a little true o f you, “three” if the statem ent is m ostly true o f you, or “four” if the statem ent is very true or you. Not at all true of me 1 A little true of me 2 Mostly true o f me 3 1. I am an honest person. 2. I care about what happens to others. 3. It’s easy for me to make up my mind about things. 4. I think I am better then most of the other people I know. 5. When someone’s feelings have been hurt, 1 try to make them feel better. 6. I can take care of myself. 7. People like me. 8. I usually speak softly. 9. I can control a lot of the kids in my class. 10. I am a serious person. 11. I am a warm person. 12. I like to do things that boys and men do. 13. I have many friends. 14. I am a kind and caring person. 15. When a decision has to be made, it’s easy for me to take a stand. 16. I usually get things done on time. 17. It is easy for people to get me to believe what they tell me. 18.1 get pretty angry if someone gets in my way. 19. It is easy for me to fit into new places. 20. Sometimes I like to do things that younger kids do. 21. I am a leader among my friends. 22. I am always losing things. 23. I don’t like to say bad words or swear. 24. I would rather do things my own way than take directions from others. 25. I am careful not to say things that will hurt someone’s feelings. 26. I like babies and small children a lot. 27. When I play games, I really like to win. 28. I like to do things that other people do. 29. I am a gentle person. 30. I am willing to work hard to get what I want. Very true of me 4 1 1 1 1 1 2 2 2 2 2 3 3 3 3 3 4 4 4 4 4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 102 Not at all true o f me 1 A little true of me 2 Mostly true o f me 3 31 .1 like to help others. 32. When there’s a disagreement, I usually give in and let others have their way. 33. I am sure o f my abilities. 34. I am a moody person. 35. I am a cheerful person. 36. I stand up for what I believe in. 37. I am the kind of person others can depend on. 38. I feel shy around new people. 39. I would rather do things on my own than ask others for help. 40. I like acting in front o f other people. 41. When I like someone, I do nice things for them to show them how I feel. 42. I am good at sports. 43. I am a happy person. 44. I feel good when people say nice things about me. 45. It is easy for me to tell people what I think, even when I know they will probably disagree with me. 46. I never know what I’m doing from one minute to the next 47. I am faithful to my friends. 48. I make a strong impression on most people I meet. 49. I always do what I say I will do. 50. I like to do things girls and women do. 51. I can get people to do what I want them to do most of the time. 52. I feel bad when other people have something I don’t have. 53. It makes me feel bad when someone else is feeling bad. 54. I like to think about and solve problems. 55. I try to tell the truth. 56. I can usually tell when someone needs help. 57. I am good at taking charge o f things. 58. I like to keep secrets. 59. I am good at understanding other people’s problems. 60. I am willing to take risks. Very true of me 4 1 1 2 2 3 3 4 4 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 1 1 1 1 2 2 2 2 3 3 3 3 4 4 4 4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 A dapted from: Boldizar, J. (1991). Assessing sex typing and androgyny in children: The Children’s Sex Role Inventory. Developmental Psychology, 27(3), 505-515. 103 Appendix B Letters for Recruitm ent o f Treatment Group Initial T etter to Clients o f Agency A Novem ber 6, 1996 D ear prospective participant: I am writing to invite you to participate in a study being conducted at [Agency A] by Kendra Rogers. Ms. Rogers is a graduate student in the M aster o f Education in Counselling program at the U niversi^ o f Northern British Columbia. Kendra has served as a practicum student and an employee at [Agency A] for the past year. She is conducting a study o f children who w itness dom estic violence and their views o f themselves as masculine or feminine. H er study w ill use questionnaires to gather inform ation from mothers and their children. This inform ation will be completely confidential, and your name w ill not be used at any time. The results o f her study w ill be available at the [Agency A] library upon com pletion o f her research project. You are welcome to read the results o f the study, as it m ay deepen your understanding o f the effects o f witnessing violence. The attached inform ation sheet will provide you w ith more inform ation about the study, which should help you in the decision whether to participate. I f you would like more information about the study, please contact M s. Rogers at : She would be glad to answer any questions you have. I f you would like to participate in the study, please telephone me or M s. Rogers to set up an appointm ent for distribution o f the questionnaires and an interview w ith you child or children. I f you do not want to participate in the study, please take a few minutes to telephone Intersect, and your name will be removed from the list. Sincerely, (signed by each prospective participant’s therapist) Agency A 104 Description o f Study: This study is an investigation o f how witnessing dom estic violence affects children’s thoughts about m asculinity and femininity. The researcher is seeking inform ation fix>m mothers regarding their children’s exposure to dom estic violence, and from children and their mothers regarding their views o f themselves as m asculine or feminine. Adult Consent for Own Participation: If you decide to participate in this study, your involvem ent w ill take approxim ately one hour. I will ask you to com plete three questionnaires. The first questionnaire gathers demographic inform ation about you and your family. The second questionnaire determines the degree to which you see yourself as masculine or feminine. The final questionnaire describes the conflict resolution styles used by you and your partner or ex-partner. It also asks about your child’s exposure to conflict between you and your partner or ex-partner. The third questionnaire asks you to think about various conflict situations you have experienced with your partner or ex-partner. Some o f the exam ples describe physical conflict between you and your partner or ex-partner. This may bring up some painful memories for you, and may cause em otional distress or anxiety. I f this situation occurs, please feel free to stop until you feel m ore comfortable. If you are unable to continue, you may discontinue participation in the study. Participation in this study may help you understand how witnessing domestic violence affects children’s view s o f men and women. Your participation in this study is completely voluntary. You m ay withdraw at any tim e throughout the study, and may refuse to answer any o f the questions. A ll inform ation you provide for this study w ill be number coded and confidential. This study has been approved by [Agency A] and die University o f Northem B ritish Columbia ethics committee. Questions? I f you have any questions about this study, contact Ms. Kendra Rogers at ###-#### or Dr. Peter M acM illan at 960-5555. Please check the appropriate box: □ □ I agree to participate in the study. I do not wish to participate in this study. I have read and understood this consent form. I understand that any inform ation about me obtained fiÿom this research will be kept strictiy confidential. Signature ____________________________ Date 105 Initial Tetter to Clients o f Apencv B Novem ber 12, 1996 Dear prospective participant: I am w riting to invite you to participate in a study being conducted at [Agency A] by Kendra Rogers. M s. Rogers is a graduate student in the M aster o f Education in Counselling program at the University o f N orthem British Columbia. Kendra has served as a practicum student and an employee at [Agency A] for the past year. She is conducting a study o f children who w itness dom estic violence and their views o f themselves as m asculine or feminine. Her study w ill use questionnaires to gather information from m others and their children. This inform ation w ill be com pletely confidential, and your name w ill not be used at any time. The results o f her study w ill be available at the [Agency A] library upon completion o f her research project. You are welcome to read the results o f the study, as it may deepen your understanding o f the effects o f witnessing violence. The attached inform ation sheet will provide you with more inform ation about the study, which should help you in the decision whether to participate. I f you w ould like more information about the study, please contact Ms. Rogers at ###-####. She would be glad to answer any questions you have. I f you w ould like to participate in the study, please telephone me or Ms. Rogers at to set up an appointm ent for distribution o f the questionnaires and an interview with you child or children. Sincerely, (signed by psychologist at Agency A) Name o f psychologist Agency A 106 F o U o w - u d Letter to Clients o f Agency A January 17,1997 D ear prospective participant: I am writing to remind you o f my recent invitation to participate in a study being conducted at [Agency ]. It is not too late to respond, as I value your input on a very im portant topic children who witness domestic violence. This study will increase our understanding o f how witnessing domestic violence affects our children. To protect your children’s well-being, they w ill not be asked any questions about witnessing dom estic violence. In fact, they will not even know that the study has anything to do with violence. The children’s questionnaire focuses completely on boys’ and girls’ personality characteristics. It is in no way threatening or harmful to the children. In order to determine the extent o f your children’s exposure to domestic violence, you will complete a questionnaire asking about specific abusive incidents between you and your partner or ex-partner. W hile it may be stressful or anxiety-producing for you to recall these incidents, you will learn a great deal about the effects o f witnessing dom estic violence on your children’s opinions o f boys and girls. All inform ation you and your children provide for this study is completely anonymous and confidential. I am the only person who will be meeting with your children, and I am the only person who will be reading the questionnaires. Your name and your children’s names will not appear anywhere on the questionnaires. As soon as the interview and questionnaires are complete, they will be number-coded. The tim e required to help w ith this study will no exceed one hour. Please take a moment to consider whether you would like to participate in this study. You can reach me at ###-#### with you decision or if you have any further questions. I would really like to hear fiom you! Sincerely, Kendra Rogers Agency A 107 Initial Letter to Clients o f Agency C January 23, 1997 Dear prospective participant: I am w riting to invite you to participate in a study being conducted at [Agency A ]. I am a graduate student in the M aster o f Education in Counselling program at the University^ o f N orthem B ritish Columbia. I have served as a practicum student and an em ployee at [Agency A] for the past year. I am conducting a study o f children who witness dom estic violence and their views o f them selves as masculine or fem inine. M y study w ill use questionnaires to gather information from m others and their children. This inform ation w ill be completely confidential, and your name w ill not be used at any tim e. The results o f my study will be available in the [Agency A] library upon com pletion o f my research project. You are welcome to read the results o f the study, as it may deepen your understanding o f the effects o f w itnessing violence. The attached information sheet w ill provide you with more inform ation about the study, w hich should help you in the decision whether to participate. If you would like more inform ation about the study, please contact me at [Agency A] at #######. I would be glad to answer any questions you have. If you w ould like to participate in the study, please telephone m e at ###-////#// to set up an appointm ent for distribution o f the questionnaires and an interview with your child or children. Sincerely, Kendra Rogers Agency A 108 Appendix C Letters for Recruitment o f Comparison Group Initial Letter to Prospective Elementary School Participants A pril 29,1997 D ear parents: A graduate student in the M aster o f Education in Counselling program at UNBC is conducting a study o f how children’s gender beliefs are affected by exposure to domestic violence. Both UNBC and School District 57 have given ethics approval for the study. The portion o f the study involving fam ilies whose children have witnessed domestic violence has been com pleted. The researcher is now looking for women whose children HAVE NOT witnessed dom estic violence. These women and their children w ill form a group which w ill be compared to the existing group o f fam ilies whose children have witnessed dom estic violence. Participation in this study is completely anonymous and confidential. Participation requirements: y your children are between the ages o f 6 and 12 years your children have not witnessed domestic violence What does participation involxe? I f you participate in this study, you w ill complete three questionnaires. The first is a dem ographic questionnaire; the second asks you to rate yourself on various personality characteristics; and the third asks about incidents o f conflict between you and your partner or ex-partner. The questionnaires take a total o f approxim ately 30 m inutes to complete. Your children w ill com plete one questionnaire which asks them to rate themselves on various personality characteristics. The questionnaire takes approximately 20 m inutes to complete. Y o u r children w ill not be asked ANY questions about dom estic violence. Benefits o f participating: If you choose to participate in this study, you w ill build our knowledge and understanding o f how dom estic violence affects our children. This information is not only valuable to those directly involved. A ll o f our lives are affected by violence, and we share the responsibility for doing som ething about it. Your children w ill each receive a sm all gift for participating in the study. 109 Please return this completed form to school with your children before May 16,1997. Y ou may keep the inform ation sheet for your records. Results o f this study w ill be available at the [Agency A] library in the fall o f 1997. Please mark the appropriate box after reading the description o f the study: □ I do not w ish to participate. □ I would like more information before I make my decision. Please phone me a t ______________ (h i o r ___________________ (w). M y name is ___________________________ . □ I would like to participate. Please phone me a t__________________ (h) o r ___________________ ( w l . My name is _______________________________. I f you have decided to participate, please list the names and birth dates o f your children who w ill be participating w ith you. Nam e Birth date I f you require more information or have any questions, please feel free to contact us. Thank you for your consideration. Sincerely, M s. Kendra Rogers or ###.# # # Dr. Peter MacM illan 960-5828 110 Follow-up Letter to Prospective Elementary School Participants Dear prospective participant: I am writing to rem ind you o f my recent invitation to participate in a study being conducted by the University o f N orthem British Columbia and School D istrict #57. I am still looking for participants for the comparison group o f children who have not witnessed any dom estic violence. These children’s questionnaires w ill be com pared to questionnaires o f a group o f children who have witnessed dom estic violence. The requirements for participation in the comparison group are: ♦ your children are betw een the ages o f 6 and 12 ♦ your children live w ith you ♦ your children have not witnessed any form o f domestic violence A ll information you and your children provide for this study is com pletely anonymous and confidential. Your name and your children’s names w ill not appear anywhere on the questionnaires. As soon as the interview and questionnaires are complete, they w ill be number-coded. The tim e required to help w ith this study will not exceed one hour. Please call me at ###-//### if you would like to participate in this study. I would really like to hear fi’om you! Sincerely, Kendra Rogers Graduate Student, University o f Northem British Columbia M aster o f Education in Counselling I ll Letter Posted nn Bulletin Board at Apencv A Dear parents; A graduate student in the M aster o f Education in Counselling program at UNBC is conducting a study o f how children’s gender beliefs are affected by exposure to domestic violence. Both UNBC and School District 57 have given ethics approval for the study. The portion o f the study involving families whose children have witnessed domestic violence has been completed. The researcher is now looking for women whose children HAVE NOT witnessed dom estic violence. These women and their children w ill form a group which w ill be compared to the existing group o f fam ilies whose children have witnessed domestic violence. Participation in this study is completely anonymous and confidentiaL Participation requirements: your children are between the ages o f 6 and 12 years ^ your children have not witnessed domestic violence What does participation involve? If you participate in this study, you will complete three questionnaires. The first is a demographic questionnaire; the second asks you to rate yourself on various personality characteristics; and the third asks about incidents o f conflict between you and your partner or ex-partner. The questionnaires take a total o f approximately 30 minutes to complete. Your children will complete one questionnaire which asks them to rate themselves on various personality characteristics. The questionnaire takes approximately 20 minutes to complete. Your children w ill not be asked ANY questions about domestic violence. Benefits o f participating: If you choose to participate in this study, you w ill build our knowledge and understanding o f how dom estic violence affects our children. This inform ation is not only valuable to those directly involved. All o f our lives are affected by violence, and we share the responsibility^ for doing something about it. Your children will each receive a small gift for participating in the study. Please leave your name and phone number w ith the receptionist if you are interested in participating in the comparison group or call Kendra at ; 112 Appendix D Instructions for Children’s Com pletion o f the CSRI I provided the following instructions to each child before h e^n n in g the CSRI: “I asked your m other to bring you here today to take part in a study o f what boys and girls think about them selves. I am going to read several sentences to you, one at a time. W hat I w ould like you to do is rate each sentence on a scale o f 1 to 4 for how well each sentence describes you (refer to poster o f scale on wall). A nsw er ‘ 1’ i f the sentence is not at all true o f you, ‘2 ’ if the sentence is a little true o f you, ‘3’ if the sentence is mostly true o f you, or ‘4 ’ i f the sentence is very true or you. Then I will w rite down your answer on this piece o f paper. Do you have any questions right now?” (If yes, answered now). “To help you get used to using these numbers for answers, let’s do three practice questions. I w ill say something, and you use these num bers to tell m e how true it is about you. The first one is, T love Brussels sprouts.’” (Child answers, w ith help if necessary). “The second one is, T would like a pet rhinoceros.’” (C hild answers, w ith help if necessary). “The last one is, T love school.’” (Child answers, w ith help if necessary). “Do you have any questions before we start?” (Answer questions as required). “Okay, let’s begin.” “As I said before, these sentences are talking about you. W hat I would like you to do is use those numbers to tell me how true it is about you. D on’t forget you can ask questions whenever you want to. Are you ready?” (Answer questions as required). Proceed through the CSRI. 113 Appendix £ Treatm ent and Comparison Group Letters o f Consent to Participate Treatment Group Consent Forms Description o f Studv: This study is an investigation o f how witnessing dom estic violence affects children’s thoughts about m asculinity and femininity. The researcher is seeking information from m others regarding their children’s exposure to dom estic violence, and from children and their m others regarding their views o f themselves as m asculine or feminine. A dult Consent for Own Participation: If you decide to participate in this study, your involvem ent will take approximately one hour. I will ask you to com plete three questionnaires. The first questionnaire gathers demographic inform ation about you and your family. The second questionnaire determines the degree to which you see yourself as masculine or feminine. The final questionnaire describes the conflict resolution styles used by you and your partner or ex-pariner. It also asks about your child’s exposure to conflict between you and your parmer or ex-parmer. The third questionnaire asks you to think about various conflict situations you have experienced w ith your partner or ex-pariner. Some o f the examples describe physical conflict between you and your partner or ex-partner. This may bring up some painful memories for you, and m ay cause emotional distress or anxiety. If this situation occurs, please feel free to stop until you feel more comfortable. If you are unable to continue, you may discontinue participation in the study. Participation in this study may help you understand how witnessing domestic violence affects children’s views o f men and women. Your participation in this study is completely voluntary. You may withdraw at any tim e throughout the study, and may refuse to answer any o f the questions. All information you provide for this study w ill be number coded and confidential. This study has been approved by [Agency A] and the University o f Northern B ritish Columbia ethics committee. Questions? If you have any questions about this study, contact Ms. Kendra Rogers at ; Dr. Peter M acM illan at 960-5555. 114 Please check the appropriate box: □ □ I agree to participate in the study. I do not w ish to participate in this study. I have read and understood this consent form. I understand that any inform ation about me obtained from this research will be kept strictly confidential. Signature ______________________ Date _____________ Parent Consent for Child’s Participation: I would like to ask your perm ission for your children to participate in a study o f dom estic violence and children’s views o f themselves as m asculine or feminine. If you agree to your children’s participation in this study, they w ill complete one questionnaire. The questionnaire assesses the degree to which the children view them selves as m asculine or feminine. Your children’s participation is completely voluntary. They may withdraw at any tim e throughout the study, and may refuse to answer any o f the questions. All inform ation obtained from your children w ill be number coded and confidential. This project has been approved by both [Agency A] and the University o f Northern B ritish Columbia’s ethics committee. Questions? If you have any questions regarding this study, please contact Ms. Kendra Rogers at ####### or Dr. Peter M acM illan at 960-5555. Please check the appropriate boxes: □ □ I give consent for my child / children to participate in this study. I do not want my child / children to participate in this study. I have read and understand this consent form. I understand that any inform ation about my children obtained from this research will be kept strictly confidential. Signature: C hild’s/Children’s Name(s): _________________________ Date_______ ___________________ 115 Child’s Consent for Own Participation: I understand that I have been asked to be in a project looking at how children think about them selves. If I % ree to be in this project, I w ill be interviewed by Kendra and asked about boys’ and girls’ personality characteristics. This will take about one hour. I understand that I do not have to answer any questions I don’t w ant to, and I can stop early if I feel uncomfortable. I f f feel bad about any o f the questions, I can talk to m y parent(s) or Kendra about it. I understand that my name w ill not be used and that Kendra w ill not tell anyone w hat I say. If I have any questions, I can ask m y parent(s) or have them call Kendra. Please check the appropriate box: □ □ I agree to participate in this project. I do not want to participate in this project. I have had the chance to ask questions. Child’s Signature Date ___________________ Interview er Signature_________________________ Date ___________________ 116 Comparison Group Consent Letters This study is an investigation o f how witnessing dom estic violence affects children’s thoughts about masculinity and femininity^. The researcher is seeking information about children’s exposure to domestic violence, and about m others’ and children’s views o f them selves as traditionally masculine or feminine. If you decide to participate in the comparison group for the study, your involvement w ill take approxim ately thirty minutes. I will ask you to com plete three questionnaires. The first questionnaire gathers dem ogr^hic information about you and your family and asks about the history o f violence against you in past relationships. The second questionnaire determines the degree to which you see yourself as masculine o r feminine. The final questionnaire describes the conflict resolution styles used by you and your partner or ex-partner. This questionnaire also asks about your child’s exposure to conflict between you and your partner or ex-partner. Some o f the examples describe physical conflict between you and your partner or ex-partner. Since you are part o f the comparison group, you should find that this questionnaire does not apply to you. However, it m ay still bring up some painful memories for you, and may cause emotional distress or anxiety. I f this situation occurs, please feel firee to stop until you feel more comfortable. If you are unable to continue, you may discontinue participation in the study. If you choose to participate in this study you w ill receive a small gift for each o f your children that participates. This study is being conducted by the University o f N orthern British Columbia and has been approved by School D istrict #57 and the UNBC ethics committee. All inform ation you and your children provide for this study will be number coded, anonymous, and confidential. Questions? If you have any questions about this study, contact Ms. Kendra Rogers at ###-////## or Dr. Peter M acM illan at 960-5828. Please check the appropriate box: □ I agree to participate in the study. □ I do not w ish to participate in this study. I have read and understood this consent form. Signature ____________________________ Date _____________ 117 Parent Consent for Child’s Participation: If you agree to your children’s participation in this study, they will complete one questionnaire each. The questionnaire assesses the degree to which the children view them selves as traditionally masculine or feminine. All information obtained from your children w ill be number coded, anonymous, and confidential. Your children’s participation is completely voluntary. I f they feel uncomfortable during the study, they may withdraw at any time, and may refuse to answer any o f the questions. Please check the appropriate boxes: □ I give consent for my child / children to participate in this study. □ I do not want my child / children to participate in this study. Childfren^’s Namefs't: ChildfrenVs Birth datefs^: M other’s Signature: ________________________________ Date Child’s Consent for Own Participation: I understand that I have been asked to be in a project looking at how children think about them selves. I f I agree to be in this project, Kendra w ill ask me a list o f questions about how I describe m yself. This w ill take about twenty minutes. I understand that my name will not be used and that Kendra w ill not tell anyone what I say. I do not have to answer any questions I don’t want to, and I can stop early if I feel uncomfortable. If I feel bad about any o f the questions, I can talk to my parent(s) or Kendra about it. If I have any questions later, I can ask my parent(s) or have them call Kendra. Please check the appropriate box: □ □ I agree to participate in this project. I do not want to participate in this project. Child’s Signature: Date ___________________ Date ___________________