Hubert Van Gijseghem, Ph.D.

psychologue, professeur émérite, Université de Montréal

“Links Between Sexual Abuse in Childhood
and Behavioural Disorders
in Adolescent Girls:
A Multivariate Approach.”

In Canadien Journal of Behavioural Science, 1994, vol. 26, no 3, 339-352.

Abstract / Résumé

Introduction

METHODOLOGY

Definition of Sexual Abuse

Subjects

Assessment instruments

Statistical analysis

RESULTS

Discrimination between samples

Parameters of sexual abuse

Effect of sexual abuse

DISCUSSION

REFERENCES

ABSTRACT

The purpose of this study is to assess the links between adolescent behaviour? (1) problems and sexual abuse in childhood. Besides a control group (N : 181), two clinical groups are tested, one presenting sexual behavioural problems (N = 62) and the other behavioural problems of a non-sexual nature (N = 107). The methodology is retrospective, using a set of questionnaires defining variables such as sexual abuse and its features, aspects of parent-child relationships, sociodemographic characteristics, etc. Multivariate analysis results show many links between behavioural disorders during adolescence and sexual abuse during childhood. Specifically, the girls having disclosed sexual abuse and belonging to the clinical groups are associated with frequent abuse, with penetration, perpetrated by a male and when the child was older, while the abuse revealed in the control group was less frequent, involved fondling only and occurred at an earlier age. Considering the clinical groups alone, sexually related behavioural disorders are associated with high frequency, severity, an adolescent abuser and a reaction of fear.

RÉSUMÉ

Cette étude cherche à évaluer les liens qui existent entre les problèmes de comportement chez l'adolescent et l'abus sexuel qu'il ou elle aurait subi au cours de l'enfance. En plus d'un groupe de contrôle (N = 181), l'étude porte sur deux groupes cliniques, dont un presente des problèmes de comportement d'ordre sexuel (N = 62) et l'autre des problèmes de comportement d'ordre non sexuel (N = 107). L'étude, retrospective, fait appel à une série de questionnaires qui définissent les variables telles l'abus sexuel et ses composantes, certains aspects des relations parent-enfant, les caracteristiques socio-démographiques, etc. Les résultats de l'analyse multivariee font ressortir de nombreux liens entre les troubles de comportement présents à l'adolescence et le fait que le sujet ait été, enfant, victime d'abus sexuel. En particulier, chez les filles des groupes cliniques qui ont revélé être victimes d'abus sexuel, on a constaté une fréquence elevée d'abus perpétré par un homme, y compris la pénétration, lorsque l'enfant était plus vieille. Par contre, l'abus revélé par les membres du groupe de contrôle était moins frequent, se limitait à des attouchements et avait eu lieu quand l'enfant était plus jeune. Pour ne considerer que les

[340]

résultats des groupes cliniques, les troubles de comportement d'ordre sexuel sont associés à une fréquence élevée des incidences, à des situations d'abus plus grave, à la présence d'un agresseur adolescent et à une reaction de peur.

INTRODUCTION

The consequences of sexual abuse suffered in childhood have been thoroughly studied in adult female populations. However, only in the last few years have researchers begun to look at the effects of sexual abuse on adolescents. Results suggest that large numbers of adolescent girls with behavioural disturbances were the victims of sexual abuse during childhood. This becomes evident with regard to behavioural problems of a sexual nature, juvenile prostitution and sexual promiscuity having been shown to be strongly related to sexual abuse (Bour, Young & Henningsen, 1984 ; Burgess, 1984 ; Fromuth, 1986 ; McMullen, 1987 ; Silbert & Pines, 1983). A number of other adolescent behavioural disorders may also be associated with sexual abuse : running away from home (Janus, Burgess & McCormack, 1987 ; Lindberg & Distad, 1985 ; Rimsza, Berg & Locke, 1988), drug addiction (Benward &c Densen-Gerber, 1975 ; Lindberg & : Distad, 1985), criminal and delinquent behaviour (Rosenthal & Doherty, 1985 ; Scott & Stone, 1986 ; Widom, 1989), self-mutilation and suicidai behaviour Qanus et al., 1987 ; Lindberg & Distad, 1985 ; Rimsza et al., 1988), depression (Orr & Downes, 1985 ; Scott & Stone, 1986 ; Sedney & Brooks, 1984) and severe identity disorders (Husain & Chapel, 1983 ; Sansonnet-Hayden, Haley, Marriage & Fine, 1987).

In addition to the study of links between sexual abuse and behavioural disorders, researchers have turned their attention to the impact of different specific variables of sexual abuse. Efforts have been made to determine why some victims are more affected by abuse than others. Despite contradictions in some of the findings, studies suggest that the severity of effects depends on the length of the period of abuse and on its frequency (Russell, 1983,1984, 1986 ; Briere, Evans, Runtz & Wall, 1988 ; Tsai, Feldman-Summers & Edgar, 1979), on the age of the child when the abuse began (Finkelhor, 1979 ; Russell, 1986), on the presence of coercion (Bagley & Ramsey, 1985 ; Briere et al., 1988 ; Elwell & Ephross, 1987), on the presence of violence (Russell, 1986 ; Finkelhor, 1979 ; Fromuth, 1986), on the degree of severity of the abuse (Russell, 1986), on the age difference between abuser and victim (Briere et al., 1988 ; Finkelhor, 1979 ; Fromuth, 1986), on the number of abusers (Finkelhor, 1984) and on the sex of the abuser (Finkelhor, 1979). Some studies focussed on the differences between intra- and extrafamily abuse, suggesting that incest has the worst effects (Briere et al., 1988 ; Gomes-Schwartz, Horowitz & Sauzier, 1985 ; Russell, 1984) and on the reaction of the child following the abuse, even if data are conflicting in the latter case (Conte & Schuerman, 1987 ; Benward & Densen-Gerber, 1975 ; Nielson, 1983 ; Tsai et al., 1979). Other contradictory findings concern variables such as the role of disclosure, the impact of family [341] support and the reaction of the mother following disclosure.

Some studies focus on the possible links between behavioural problems, sexual abuse and different family variables. A number of family variables have been identified as definite risk factors of sexual abuse (v.g., Gruber & Jones, 1983). On the other hand, most authors find it hard to differentiate between the consequences of the abuse per se and those resulting from underlying family dysfunction (v.g., Gruber & Jones, 1981 ; Fromuth, 1986). Only sophisticated statistical procedures can provide answers in this area.

Doubtless, a vast number of variables determine the diversity and relative severity of the consequences of sexual abuse, whether these variables are abuse-related or belong to the underlying family characteristics. However, research in this area is necessarily founded on the hope of developing a scale of severity that would predict possible consequences. Only with such a scale could one target specific prevention efforts. The present study intends to be a further step in that direction. Its purpose is to assess the links between certain adolescent behavioural disorders and sexual abuse in its specific features and also to take into account a number of family variables. Special attention was given to disorders of a sexual nature in an effort to determine if abuse had a particular impact on the sexual aspect of adolescent behaviour.

Three main hypotheses were put forward : first, that adolescents with behavioural disorders requiring psychosocial intervention had experienced more sexual abuse than adolescents without behavioural disorders ; second, that adolescents with behavioural disorders of a sexual nature requiring psychosocial intervention had experienced more sexual abuse than adolescents with behavioural disorders of a non sexual nature ; and third, that adolescents who had experienced sexual abuse and who exhibited behavioural disorders (of a sexual or non sexual nature) had experienced a more severe degree of abuse in its different features than adolescents who had experienced sexual abuse, but who did not exhibit behavioural disorders. In addition to the main hypotheses, a number of questions were raised regarding the impact of family variables, especially parent-child relationships, on the consequences of sexual abuse.

METHODOLOGY

Definition of Sexual Abuse

The authors relied on the work of Finkelhor (1979, 1984), Meiselman (1978) and Van Gijseghem (1988) to arrive at the following definition :

Sexual abuse is the exposure of a child to sexual stimulation, involving full or partial hetero or homosexual intercourse, as well as sexual games for the purpose of sexual stimulation of the child or of the perpetrator him or herself. Therefore, sexual abuse can mean coitus, masturbation, sexual touching and fondling, sexualized kissing, or acts of indecent exposure. It involves a child of up to 13 years of age and a partner 5 years older or more.

[342]

Subjects

Three hundred and fifty adolescent girls between the ages of 13 and 19 (mean age of 16) agreed to participate in the study. Of these, 107 presented behavioural disorders of a non-sexual nature which had required psychosocial intervention (clinical group #1), 62 showed sexually-related behavioural disorders which had required psychosocial intervention (clinical group #2), and 181 presented no problems requiring psychosocial intervention (control group). The girls of the control group were chosen at random from a high school in Montreal. The girls in the clinical groups lived in rehabilitation centers at the time of the study, and were divided into the two samples in accordance with the information contained in their files. The files usually included a report from the Quebec Youth Protection Tribunal, social and psychological evaluations, and reports from the rehabilitation centre personnel. In cases where a file indicated juvenile prostitution or sexual promiscuity, the subject was placed in clinical group #2 (sexual behavioural disorders). In cases where the file mentioned problems such as truancy, theft, incorrigibility (non-sexual) or being beyond control of parent or guardian, the subject was placed in clinical group #1 (behavioural disorders). In order to protect the privacy of the subjects, the person responsible for the data gathering used a code to identify the subjects according to their respective groups. While the strictest standards were respected in forming the clinical groups, they nonetheless remained a non-standardized and subjective measure, subject to expert judgement. As a result, it is imperative to proceed with care in the interpretation of the differences between the two clinical groups. It should be noted, however, that the size of the samples was, to a certain degree, a compensating factor.

Assessment instruments

Three self-report questionnaires were used to measure sexual abuse, its various specific features, aspects of the parent-child relationship, socio-demographic characteristics, covert delinquency, the presence of non-abusive sexual activity (sexual exploration between children) and observation of parental intercourse during childhood (henceforth called "primal scene"). [1]

The variables measuring prevalence and specific features of sexual abuse were drawn from a questionnaire prepared by Finkelhor (1979) in his study of a New-England college student population. This is a self-report consisting of objective questions and optional answers. In addition to the prevalence of sexual abuse, this questionnaire defined 14 variables describing the specific circumstances of sexual abuse. The variables included : age of child when abuse began, age and gender of abuser, relationship between the two, whether living under the same roof, coercion, duration and frequency, degree [343] of severity of the abuse (measured by the perpetrator's acts), presence of penetration of one of the child's body orifices (mouth, vagina and/or anus), child's reaction (fear, pleasure, curiosity), disclosure of the abuse and abuse by more than one person.

The parent-child relationship was measured through the use of the Parental Bonding Instrument (Parker, Tupling & Brown, 1979). It measures the perceived quality of parenting received from each parent and the degree of control exercised by the parents over the subject during childhood. It is a subjective and retrospective measure. The four scales of this instrument were calculated with principal component analysis and they, therefore, remain empirical and of normal distribution. The reliability and validity of this instrument are satisfactory : an internal consistency coefficient of 0.70, a test-retest coefficient of 0.76 for the care factor and of 0.63 for the control factor, a split-half reliability of 0.88 for care and 0.74 for control. Intercorrelations of 0.85 and 0.69 were obtained with the opinion of independent clinicians. The concomitant validity with accompanying interview was estimated at 0.78 for the care and at 0.50 for the control.

The sociodemographic variables used in this study were : present age of the subject, type of family in which she grew up (nuclear, extended, single-parent, reconstituted family, institution or group home, etc.), size of family and socio-economic status of parents. These variables were drawn from questions integrated in the Finkelhor questionnaire.

The questionnaires were distributed among the three groups on a voluntary and anonymous basis. The rate of participation was 100% in the control group and 79.8% in the clinical groups. Subjects with minor reading problems were assisted by an adult. Two subjects with major reading problems were excluded. In view of the total confidentiality of the questionnaires, it was not possible to effect any follow-up on subjects having revealed sexual abuse.

Statistical analysis

A stepwise discriminant analysis (backward direction) was used to test the hypotheses. The first analysis, aimed at testing the first two hypotheses, was used on all subjects and retained the variables discriminating the three samples. The second analysis, aimed at testing the third hypothesis involved subjects having revealed sexual abuse, and it retained the variables discriminating the three sub-groups of abused girls. Given the fact that there exists no single rule governing recourse to selective analysis, the standardized discriminant scores were used to extract the irrelevant variables from the model until a coherent and statistically significant model was obtained (Klecka, 1980).

Some discriminant variables did not respond to the assumptions of the analysis model (normal distribution and equal group co-variance matrices). However, in view of the size of the samples, these deficiencies were largely [344] compensated. In addition, it is recommended that in such cases the between group variance rather than the test of significance be used, and that there be a high percentage of correct classification (Klecka, 1980).

RESULTS

The rate of disclosure of sexual abuse was 46% (50/107) in clinical group #1 (behavioural disorders), 68% (42/62) in clinical group #2 (sexual behavioural disorders) and 11% (20/181) in the control group. These results indicate a much higher prevalence of sexual abuse in clinical groups # 1 and # 2 than in the control group (p < .001) and also a higher prevalence of sexual abuse in clinical group # 2 than in clinical group # 1. The first two hypotheses are therefore confirmed.

The child's age at the onset of abuse varied from one to 13 years, with a mean age of 8 ; the abuser's mean age was 28. Nearly 30% of abusers were under 18, and 91% were male ; in 33% of the cases the abuser was a parent or. step-parent ; in another 27% he or she was family-related ; in 34% of the cases he or she was known but unrelated to the victim and in 7% of the cases he or she was a stranger. In 43% of the cases the abused child lived with the perpetrator. Nearly 42% of the respondents reported a high degree of coercion (threats, use of force), while 35% reported they were coerced by promises and 23% said they had experienced little coercion. The abuse lasted less than one year in 44% of the cases, between one and two years in 25% of the cases and more than three years in 31% of the cases. The abuse occurred only once in 33% of the cases. Respondents reported a reaction of fear, repulsion and disgust in 33% of the cases, and a reaction of pleasure in 29% of cases. Others reported a mixed reaction. In 22% of the cases the child never [345] revealed her experience to anyone. Finally, 29% of the abused girls reported having experienced additional sexual abuse.

Discrimination between samples

The findings of discriminant analysis used for all the girls and the position of the three samples throughout the length of the two retained functions are shown in Table 1 and Figure 1. Function 1 (91% of between group variance, p < .0001) defined the variables distinguishing the control group from the clinical groups, while Function 2 (9% of between group variance, p < .001) defined the variables distinguishing the two clinical groups. Together, they correctly classified 68% of all cases : 84% (152/181) of the control group, 46% (49/107) of clinical group #1 and 62% (38/62) of clinical group #2. Given that left to chance, the correct classification would have been 33%, these rates are very high and ensure the validity of the findings.

In addition to a higher prevalence of sexual abuse in clinical groups #1 and #2, the type of family in which the subject grew up also determined to a high degree the sample to which she belonged (score of ± 0.64 on Function 1 and ± 0.37 on Function 2). Thus, subjects coming from an unstable family (extended, single-parent or reconstitued family, group home or institution) were more likely to belong to the clinical groups, particularly to clinical group #2, than the subjects from a stable, nuclear family. In addition, the girls of the control group seemed to have received better maternal care (0.32), to have had a higher socio-economic status (0.28) and were slightly older at the time of the investigation (0.22) than the clinical group girls.

While showing a higher prevalence of sexual abuse, the girls in clinical group 42 were older (0.63) at the time of the investigation than those of clinical group #1 ; they also had experienced a higher level of maternal control [346] (0.39), a lower level of parental control (-0.32) and a lower rate of sexual experience among peers (-0.33). They came from a more stable (0.37) but smaller-sized family (-0.25) and had experienced more observation of primal scene than the girls of clinical group #1 (0.25).

Parameters of sexual abuse

The discriminant analysis findings applied to the girls who had reported experiencing sexual abuse during childhood, as well as the position of the samples throughout the length of the two retained functions, are shown in Table 2 and Figure 2. Function 1 (71% of between group variance, p < .0001) distinguished the control group from the clinical groups, while Function 2 (29% of between group variance, p < .01) distinguished the clinical groups. Together, they correctly classified 69% of the cases : 84% (17/20) of the control group, 64% (32/50) of clinical group # 1 and 66% (28/42) of clinical group # 2. It should be noted again that these are relatively high percentages compared to what would have been obtained by chance alone.

The discriminant analysis of the girls reporting abuse revealed the following differences between the clinical and the control groups. The clinical group girls were distinguised by frequent abuse (0.47), with penetration (0.36), perpetrated by a male (0.37) and when the child was older (0.45). The control group girls were associated with less frequent abuse, involving fondling only, at an earlier age and more often perpetrated by a female. The third hypothesis is therefore confirmed.

[347]

Considering the clinical groups alone, the abused girls in group # 2 (sexual behavioural problems) were more likely to have been abused by an adolescent (-0.52) and to have had a reaction of fear (0.51). Within this group the abuse was frequent (030) and severe (0.30). Also, compared to clinical group tt 1, the girls of group # 2 were older in age at the rime of the investigation (0.43). Finally group #2 girls had, in more than one occasion, been witness to primal scene (0.29).

With regard to family variables, the abused girls of the clinical groups were associated with an unstable and non-nuclear family (-0.64) while the abused girls of the control group came from a stable and nuclear family. Also, the clinical groups were differentiated from the control group by poor care (-0.50) and weak maternal control (-0.42). Considering the clinical groups alone, the girls of group # 2 came from a slightly more stable (0.33) and smaller-sized family (-0.43), in which maternal care was of poor quality (-0.24), maternal control was higher (0.21) and paternal control was lower (-0.40).

Effect of sexual abuse

A third discriminant analysis was conducted using sexual abuse as a group variable. Since only two groups are considered (abuse or not) one significant function was found (p < .001), leading to 76% of all the cases correctly classified. The findings of this discriminant analysis are shown in Table 3.

It appears that abused girls are more likely to develop behavioural problems (0.81). They also are differentiated from the non-abused girls by a higher paternal control (0.33) and a lower maternal control (-0.27). Other variables are neglectable.

[348]

DISCUSSION

The results indicate that adolescent girls with behavioural disorders, particularly of a sexual nature, are more likely to have experienced sexual abuse during childhood than adolescent girls without behavioural disorders. These findings give significant support to the hypothesis that sexual abuse is associated with disturbances in the girl's development.

First, discussion will deal with some abuse variables that impact on adolescent behaviour, as well as others that do not. Next, family variables will be considered. Finally, some thoughts will be devoted to the association between the sexual nature of the abuse and specific problems in adolescence.

When comparing abused girls of the control group and the clinical groups, the results clearly indicate that behavioural disorders are particularly associated with frequent abuse, a high degree of contact including penetration, and a male abuser. These findings were expected to distinguish between groups and they concur with existing research (Russell, 1984, 1986 ; Tsai, Feldman-Summers & Edgar, 1979 ; Draijer, 1988). Other findings are more confusing. For instance, the fact that the control group girls were abused much earlier in life than the clinical group girls is surprising in the light of previous data on this topic (Bigras, 1987 ; Meiselman, 1978 ; Russell, 1986). Could it be that the meaning of the abuse escapes the younger child so that the consequences are less? Another possible explanation is that the clinical group girls suffered from ongoing abuse without recall of an early onset, whereas the control group girls recalled the early abuse because the abusive relationship stopped at a given rime.

It is interesting to note that the relationship between abuser and abused does not distinguish between the abused girls of the different groups. In regard to this relationship, in 91% of the cases the abuser was well known by [349] the child but the outcome of the abuse was identical regardless of the status of the abuser, be he a father, a brother, an uncle, a neighbour or a babysitter. This finding contradicts a number of observations according to which intrafamily abuse, or incest, is considered qualitatively different from extrafamily abuse, at least with regard to its consequences (Briere et al., 1988 ; Gomes-Schwartz et al., 1985 ; Russell, 1984). The data obtained in this study suggest that the significance of the abuser is more important to the child than his actual status.

The presence or absence of overt coercion is another variable that fails to distinguish between the different groups of abused girls. This finding gives some weight to the assertions of Berliner & Conte (1990) and German, Habenicht & Futcher (1990) that the definition of coercion should be extended to include seduction, promises or gifts as well as physical or moral violence. Distinctions between different forms of coercion could therefore be thin.

When comparing abused girls of the two clinical groups, one observes that the abuser of clinical group #1 girls (behavioural problems) is significantly older than the abuser of clinical group #2 girls (sexual behavioural problems). In fact, the latter is likely to be an adolescent. Again this finding is unexpected since results of other studies suggest the opposite (v.g., Briere et al., 1988 ; Fromuth, 1986). Also, at least one theoretical model suggests that the consequences of sexual abuse are more severe (or affect the child's identity, including sexual) if the abuse means the abolishment of the distance between generations (Van Gijseghem, 1985). If this theory has some validity, the present findings could indicate that the generation gap is not necessarily a function of age but one of difference in developmental stage and experience. The fact that the clinical group #2 girls seem to be more disturbed, even if the abuser is younger, could also be explained by the absence, in the attitude of the younger abuser, of progressive seduction or tenderness. This same hypothesis could also explain our finding that the clinical group #2 girls reacted negatively to the abuse (fear, disgust) compared with the clinical group #1 girls, whose reactions were more positive.

With regard to the family factors that play a role in the impact of sexual abuse, one notices first of all that, unlike the abused control group girls, the abused girls with behavioural disorders (sexual and non-sexual) tend to come from unstable families and have experienced poor maternal affection and control. In fact, the variable of maternal and paternal control shows up in a persistent way throughout our results. This variable, generated by Parker's Parental Bonding Instrument, indicates whether the adolescent, as a child, experienced too much or too little parental (mother's or father's) control. The variable first shows up as discriminating significantly between abused and non-abused girls, regardless of the group to which they belonged. It appears that abused girls experienced high paternal control but lacked maternal control, whereas the opposite was true in the case of non-abused girls (high maternal control and low paternal control). These results concur with [350] previous research. Both Meiselman (1978) and Finkelhor (1984) found that, characteristically, abused girls came from families with conservative (if not patriarcal) values. Other data in Finkelhor's study indicate that girls living without their natural mother are three rimes more vulnerable to sexual abuse than the average girl. He concludes that maternal control is of paramount importance in the protection of daughters. Our data confirm this in an unequivocal manner.

With regard to the possible association between sexual behavioural problems and the sexual nature of the abuse, our findings support the assumption presented by Finkelhor & Browne (1985) that a particularly traumatic sexual experience may lead to disorders of a sexual nature, whereas other factors may lead to disturbances in other areas of development. These authors point out, however, that while this association has received strong empirical support with regard to adult women who were abused in childhood, it still needs to be established in the case of adolescents. In the light of our own results, with the rate of sexual abuse at 68% in the group of girls with sexually related behaviour disturbances, it would seem that, as significant as non-sexual factors may be, the sexual nature of the abuse in girls plays a major role in their future development.

In sum, this study's findings strongly suggest that the severity of sexual abuse, when accompanied by poor family support, jeopardizes psycho-social adjustment in adolescence. Even if abused girls in the control group are free of behavioural symptoms, the abuse having been less severe and family support of a better quality in their cases, it is still too early to predict whether these girls will develop into well-adjusted adults.

This research srudy was made possible thanks to a grant from the Conseil québécois de la recherche sociale (No. RS - 1163 M 187). Requests for reprints should be sent to Hubert Van Gijseghem, 80 Somerville, Montreal, (Canada), H3L 1A2.

REFERENCES

Bagley, C, & Ramsey, R. (1985). Disrupted childhood and vulnerability to sexual assaults : Long term sequels with implications to counseling. In C.A. Courtois (1988). Healing the incest wound : Adult survivors in therapy. New-York : Norton.

Benward, J., & Densen-Gerber, J. (1975). Incest as a causative factor in antisocial behaviour : An exploratory study. Contemporary Drug Problems, 4, 323-346.

Berliner, L., & Conte, J.R. (1990). The process of victimization : the victim's perspective. Child Abuse and Neglect, 24, 29-40.

Bigras, E. (1987). Le refus de n'être qu'un. Topique, 32, 61-78.

Bour, D.S., Young, J.P., & Henningsen, R. (1984). A comparison of delinquent prostitutes and delinquent non-prostitutes on self concept, Journal of [351] Offender Counseling Services and Rehabilitation, 9 (1-2), 89-101.

Briere, J., Evans, D., Runtz, M., & Wall, T. (1988). Symptomatology in men who were molested as children : A comparison study. American Journal of Orthopsychiatry, 58, 457-161.

Burgess, A. (Eds.) (1984). Child pornography and sex rings. Lexington : Lexington Books.

Conte, J.R. (1985). The effects of sexual abuse on children : A critique and suggestions for future research. Victimology : An Internaltional Journal, 10 (1-4), 110-130.

Conte, J.R., Se Schuerman, J.R. (1987). Factors associated with an increased impact of child sexual abuse. Child Abuse and Neglect, 11, 201-211.

Draijer, N. (1988). Een lege plek in mijn geheugen. Seksueel misbruik van meisjes door verwanten, (A blank in my memory. Sexual abuse of girls by family members). The Hague : Ministry of Social Services.

Elwell, M.E., & Ephross, P.H. (1987). Initial reactions of sexually abused children. Social Casework, 68, 109-116.

Finkelhor, D. (1979). Sexualy victimized children. New-York : Free Press. Finkelhor, D. (1984). Child sexual abuse. New theory and research. New-York : Free Press.

Finkelhor, D., Se Browne, A. (1985). The traumatic impact of child sexual abuse : A conceptualization. American Journal of Orthopsychiatry, 55 (4), 530-541.

Fromuth, M.E. (1986). The relationships of childhood sexual abuse with later psychological adjustment in a sample of college women. Child Abuse and Neglect, 10, 5-15.

German, D., Habenicht, J., & Futcher, W. (1990). Psychological profile of the femal adolescent incest victim. Child Abuse and Neglect, 14, 429-438.

Gomes-Schwartz, B., Horowitz, J.M., & Sauzier, M. (1985). Severity of emotional distress among sexually abused preschool, school-age, and adolescent children. Hospital and Community Psychiatry, 36, 503-508.

Gruber, K.J. Se Jones, R.J. (1981). Does sexual abuse lead to delinquent behavior? A critical look at the evidence. Victimology : An International Journal, 6, 85-91.

Gruber, K.J., Se Jones, R.J. (1983). Identifying determinants of risk of sexual victimization of youth : A multivariate approach. Child Abuse and Neglect, 7, 17-24.

Husain, A., & Chapel, J.L. (1983). History of incest in girls admitted to a psychiatric hospital. American Journal of Psychiatry, 140 (5), 591-593.

Janus, M.D., Burgess A.W., & McCormack, A. (1987). Histories of sexual abuse in adolescent male runaways. Adolescence, 22 (86), 405-417.

Klecka, W.R. (1980). Discriminant analysis. Beverly Hills : Sage University paper.

Lindberg, F.H. & Distad, L.J. (1985). Survival responses to incest : Adolescents in crisis. Child Abuse and Neglect, 9, 521-526.

McMullen, R.J. (1987). Youth prostitution : A balance of power. Journal of Adolescence, 10, 35-43.

[352]

Meiselman, K. (1978). Incest. A psychological study 0/ causes and effects with treatment recommendations. San-Francisco : Jossey-Bass.

Nielson, T. (1983). Sexual abuse of boys : Current perspectives. Personnel and Guidance Journal, November, 139-142.

Orr, D.P., & Downes, M.C. (1985). Self-concept of adolescent sexual abuse  victims. Journal of Youth and Adolescence, 14(5), 401-410.

Parker, G., Tupling, H., & Brown, L.B. (1979). A parental bonding instrument. British Journal of Medical Psychology, 52, 1-10.

Rimsza, M.E., Berg, R.A., & Locke, C. (1988). Sexual abuse : Somatic and emotional reactions. Child Abuse and Neglect, 12, 201-208.

Rosenthal, P.A., & Doherty, M.B. (1985). Psychodynamics of delinquent girl's rage and violence directed toward mother. Adolescent Psychiatry, 12, 281-289.

Russell, D. (1983). The incidence and prevalence of intrafamilial and extra-familial sexual abuse of female children. Child Abuse and Neglect, 7, 147-154.

Russell, D. (1984). The prevalence and seriousness of incestious abuse : Stepfathers versus biological fathers. Child Abuse and Neglect, 8,15-22.

Russell, D. (1986). The secret trauma : Incest in the lives of girls and women. New-York : Basic Books.

Sansonnet-Hayden, H., Haley, G., Marriage, K., & Fine, S. (1987). Sexual abuse and psychopathology in hospitalized adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 26(5), 753-757.

Scott, R., & Stone, D. (1986). MMPI measures of psychological disturbance in adolescent and adult victims of father-daughter incest. Journal of Clinical Psychology, 42, 251-259.

Sedney, M.A., & Brooks, B. (1984). Factors associated with a history of childhood sexual experience in a nonclinical female population. Journal of the American Academy of Child and Adolescent Psychiatry, 23(2), 215-218.

Silbert, M.H., & Pines, A.M. (1983). Early sexual exploitation as an influence in prostitution. Social Work, 28(4), 285-289.

Tsaï, M., Feldman-Summers, S., & Edgar, M. (1979). Childhood molestation : Variables related to differential impacts on psychosexual functioning in adult women. Journal of Abnormal Psychology, 88, 407-417.

Van Gijseghem, H. (1985). Autre regard sur les conséquences de l'inceste père-fille. Revue canadienne de Psycho-Éducation, 11, 73-77.

Van Gijseghem, H. (1988). La personnalité de l'abuseur sexuel. Montreal : Meridien. Widom, C.P. (1989). The cycle of violence. Science, 244, 160-166.

Submission February 28, 1993

Revision May 27, 1993

Acceptance July 13, 1993