Psychometric Assessment of the Burgess-Partner Abuse Scale for Teens (B-PAST)

Stephanie E. Burgess* Abbas Tavakoli**

* PhD, APRN-BC. Clinical Professor, College of Nursing , University of South Carolina, Columbia. scburgess@aol.com ** Dr.PH. Director of Statistics Laboratory, College of Nursing, University of South Carolina, Columbia, SC 29208. abbas.tavakoli@sc.edu

ABSTRACT

Purpose. The purpose of this study was to assess the psychometric properties of the Burgess-Partner Abuse Scale for Teens (B-PAST) by examining internal consistency and construct validity.

Methods. A convenience sample of 239 teen girls ages 13 to 19 years old that were seen in three private primary health care practices agreed to participate. Subjects were administered two measures of interpersonal violence, two measures of psychosocial well-being, and a demographic questionnaire. Construct validity was assessed by 1) conducting exploratory factor analysis with a priori decision to retain two factors, and 2) measuring differential correlates of interpersonal violence and psychosocial well-being using Pearson correlation. In order to determine the instrument's ability to discriminate between teen girls who reported partner abuse and those who did not report partner abuse, differences in contrasted group means (abused vs. non-abused) were examined. One-way ANOVA was used to determine differences in interpersonal violence and psychosocial wellbeing. The internal consistency of the B-PAST was examined by setting a criterion for the alpha coefficient above .70.

Results. The results showed that 70.9% of the variance was explained by the B-PAST. Factor analysis demonstrated a 22-item scale with two distinct subscales. The B-PAST correlated with one measure of interpersonal peer violence (r = .39) and two measures of poor psychosocial well-being, low level of hope and increased depression (r = -.42, r = .46). Differences in contrasted means showed that teen girls who were abused by a partner reported higher scores for interpersonal violence. One-way ANOVA showed that abused teen girls scored higher for interpersonal violence and depression but lower in hope. The internal consistency for the final total scale was .97 and .95 for physical/sexual abuse and .91 for the social/emotional abuse subscales.

Conclusions. Results provide support for reliability and validity for the BPAST as a measure of partner abuse in teen girls ages 13 to 19.

KEY WORDS: Partner abuse, teenagers, psychometric assessment.

Propósito. Evaluar las propiedades psicométricas de la Escala para adolescentes "Burgués-Partner Abuse Scale" (B-PAST), examinando la consistencia interna y validándola.

Métodos. A una muestra de 239 niñas adolescentes, entre 13 y 19 años, que formaron parte de la práctica privada y que aceptaron participar, se le aplicó dos medidas de violencia interpersonal, dos de bienestar psicoló- gico y un cuestionario demográfico. La validez fue valorada por 1) un factor de análisis de conducta exploratoria con una decisión a priori para retener dos factores, y 2) medición de la correlación diferencial de la violencia interpersonal y el bienestar psicosocial usando la correlación de Pearson. Con el fin de determinar la habilidad del instrumento para discriminar entre niñas que informaron abuso de sus compañeros y aquéllas que no lo informaron, se examinaron las diferencias de la media de los grupos (abusadas vs. no abusadas). Un ANOVA de una vía se utilizó para determinar las diferencias de la violencia interpersonal y el bienestar psicosocial. Se examinó la consistencia interna del B-PAST, estableciendo un criterio para el coeficiente alfa sobre .70.

Resultados. El 70.9% de la variable se explicó por medio del B-PAST. El factor de análisis demostró una escala de 22 ítemes con dos subescalas diferentes. El B-PAST correlacionado con una medida de violencia interpersonal de los compañeros (r=.39) y dos medidas de bienestar psicosocial pobre, bajo nivel de esperanza y aumento de la depresión (r=-.42, r=.46). Las diferencias en los contrastes de la media mostraron que las niñas adolescentes que habían sido abusadas por un compañero reportaron cifras más altas de violencia interpersonal. El ANOVA de una vía mostró que las niñas tenían cifras más altas para violencia interpersonal y depresión y bajas para esperanza. La consistencia interna para la escala total final fue de .97 y .95 para abuso físico/sexual y .91 para las subescalas de abuso social/emocional.

Conclusión. Los resultados respaldan la confiabilidad y validez del B-PAST como una medida para el abuso entre compañeros en niñas adolescentes entre los 13 y 19 años.

PALABRAS CLAVES: Abuso de compañeros, adolescentes, evaluación psicométrica

Partner abuse of teen girls is recognized as an emerging health problem in the United States (1). Studies suggest an incidence rate of 21% to 60% in teen girls (2-13) and 10% to 22% in pregnant teen girls (14-21). Likewise, other studies report that annually 6% of teen girls are murdered by partners and 8% are sexually assaulted by partners (22-28). Problems reported with teen partner abuse include depression (13, 29-30), less hope (31-34), peer violence/ harassment (15, 28, 35-36) and substance abuse (8, 12-13, 16, 37-39). Despite these findings, preventing teen partner abuse remains difficult for a myriad of reasons including the lack of age appropriate screening measurement scales (40-43). This gap is heightened by the U.S Preventive Services Task Force (USPSTF) 2004 findings that suggested a void in scales designed to measure partner abuse in different populations and age groups (42). According to the Task Force, current partner abuse scales (44-48), although widely used with extensive reliability and validity reports, are not linked with outcomes, define partner abuse differently, were developed and tested on adult women living in shelters or with a male partner, and depending on the scale, are limited in their measurement of social or sexual partner abuse except for forced sex or rape (42).

Literature Review

It has been suggested that teen partner relationships differ from adult partner relationships (3-25, 49-52), especially in social and sexual aspects, therefore rendering adult partner abuse instruments possibly invalid or unreliable for use in teen girls. In order to determine if current instruments are appropriate for measuring partner abuse in teen girls, a critique of each was conducted.

The Index of Spouse Abuse (ISA) is a 30-item 5-point Likert type scale designed to measure the magnitude or severity of both physical and nonphysical partner abuse of adult women in heterosexual relationships (46). The ISA was initially evaluated in samples of women living in protective shelters and college aged or adult married women living with a male partner. The initial internal consistency ranged from .90 to .94 for the physical subscale and .91 to .97 for the non-physical subscale (46). The ISA was shown to discriminate between women who were abused vs. those who were not abused in that abused women scored higher for partner abuse and personal and social problems (e.g., depression, low selfesteem, sexual dissatisfaction, and family discord) (46).

The original ISA has been revised resulting in two separate scales: Partner Abuse Scale-Physical (PAS-P) and the Partner Abuse Scale-Non-Physical (PAS-NP) (45). The two scales reflect partner abuse in nontraditional couples containing three items relating to sexual and four items relating to social partner abuse (45). Tested among 90 adult women living in shelters or with partners, the PASP and the PAS-NP produced internal consistency reliability coefficients of .90 to .95, respectively (45).

The Severity of Violence Against Women Scale (SVAWS) is a 4- point Likert type scale consisting of 46 items that were constructed to measure the magnitude (severity) of perceived threats, attempts, or completed physical violence acts on adult women by male partners in an intimate relationship (48). Tested in adult women, the initial internal consistency was .93 (48).

Based on Orem's Theory of Self-Care (53), the Danger Assessment instrument was conceptualized by Campbell to help battered women enhance their personal awareness of danger for homicide from a partner (44). Tested among 79 adult battered women who were recruited from shelters, the Cronbach alpha was .84 (44). Support for construct validity was provided with a moderate to strong correlation (r = .43) with the Conflict Tactics Scale (44, 47). Conceptualized as a one-dimensional instrument, the Danger Assessment consists of 15 dichotomous items (yes/no). Fourteen items relate to physical danger, including forced sex, and one item relates to a partner's attempt to control household money and the car (44).

The Conflicts Tactics Scale (CTS) was developed to assess a married couple's conflict in three domains: reasoning, verbal aggression, and physical aggression (47). The scale originated from the literature in the 1970s on victimization, wife beating, marital violence, and wife abuse (47). The CTS consists of physical abuse items (e.g., "does he hit, push, kick, bite, or shove you?"), emotional items (e.g., "does he threaten, insult, or swear at you?"), and reasoning ("does he make you cry?") (47). The initial internal consistency reliability was .79 among married women with followup studies producing internal consistencies of .80 to .83 among married and non-married heterosexual couples(47).

In contrast to items measured by adult partner abuse scales, the substantive literature review showed that teen partners typically used more sexual and social control of teen girls (13-25, 51). Tactics included intimidation, verbal harassment, humiliation, and using drugs/alcohol to diminish the teen girl's awareness to protect herself in sexual activity (13-25, 51). Partners also dictated what clothes to wear or what social activities the teen girls could participate in. Partners controlled whom the teen girls could associate with, talk to, or hang around with, especially at school or at social events (4-12). Additionally abusive partners demanded that teen girls account for their whereabouts, time and activities by reporting in by cell phone or pager, so that the partners could find them at a moment's notice (52).

In summary, important methodological issues emerge when adult scales are used to measure teen partner abuse. First, adult partner abuse instruments were developed and tested for use among adult women, rendering validity and reliability concerns for their use in teen girls. Second, adult partner abuse instruments are limited, if of any use, in their measurement of social or sexual partner abuse. Third, none measure social or sexual aspects specific to teen relationships. The Burgess-Partner Abuse Scale (B-PAST) is intended to measure partner abuse in teen girls, paying particular attention to aspects of social and sexual teen partner abuse (52). According to experts, an instrument that can measure teen partner abuse will enhance the opportunity to identify correlates, confirm risk factors, develop interventions, and validate outcomes that indicate that assessing teen girls for partner abuse prevents or reduces disability (42). The major purpose of this study was to assess the psychometric properties of the B-PAST. It was hypothesized that the B-PAST 1) would be composed of two separate but related teen partner abuse dimensions, 2) would positively correlate with peer violence and depression but negatively with hope, and 3) would report higher scores for interpersonal violence and depression but less hope in abused teen girls.

Conceptual Framework

Scholarly attention to partner abuse has increased in the past 30 years in order to offer explanations about factors that create and perpetuate partner abuse (54-55). For this study, an ecological framework was used, which explains that partner abuse is in part the result of interplay between situational and personal factors (56-59). Situational factors are defined as perceived or actual interpersonal interactions between individuals. Interactions can encompass interpersonal violence with a partner or peers. Personal factors are defined as demographic characteristics and psychosocial well-being. Psychosocial well-being is often influenced by interactions with individuals and with the larger community. For this study, the B-PAST was assessed by measuring correlations between the interpersonal interactions of the teen girl, her partner, and peers. Other differential correlates were assessed between the teen girl's level of hope and depression, and partner abuse. In order to determine the instrument's ability to discriminate between teen girls who reported partner abuse and those who did not report partner abuse, differences in contrasted group means (abused vs. non-abused) were examined for abuse. One-way ANOVA was used to determine differences in interpersonal violence, hope, and depression.

Methods

Subjects

A convenience sample composed of 257 teen girls ages 13 to 19 agreed to participate. Criteria for inclusion for this study were teen girls: 1) ages 13 to 19, 2) currently enrolled in school either part time or full time, 3) reporting a current partner (past 12 months), and 4) able to read and write English. After 18 surveys were removed for failing to meet study criteria, the final sample (n = 239) was composed of 62.76% (n = 150) Caucasian and 35.98% (n = 86) African-American. The mean age of the teen girl was 16.50 years with a mean education of 11.27 years. The mean age of the partner was 19.03 years with a mean education of 9.9 years (Table 1). Forty-three percent (n = 103) reported partner abuse based on a B-PAST score of 9/88. No subjects reported a same-sex partner. The mean age for teen girls initiating sexual activity was 14.26 years, and 44% (n = 105) reported 3 to 6 lifetime sexual partners. Sixteen percent (n = 38) reported a current pregnancy. Twelve percent (n = 27) reported they had witnessed their parents hitting each other. Twenty-three percent (n = 55) reported they used drugs (marijuana, ecstasy, or cocaine) and 26.1% (n = 63) reported their partners used drugs (marijuana, ecstasy, or cocaine). Forty-two percent (n = 101) were covered by the state's government health insurance plan (Medicaid) issued to low-income individuals and 57.4% (n = 137) were covered by private health insurance plans.

Procedure

Data collection occurred over an eight-month period following Institutional Review Board approval. Prior to data collection, the investigator received in-depth training to ensure uniformity to a written data collection procedure, protecting human subjects, maintaining confidentiality, avoiding study coercion, and solving ethical situations for teens seeking help from abusive partner relationships. Three private health care sites agreed to provide access to potential participants for data collection. Subjects were recruited and invited to participate in the study after they had completed their health care visit. Informed consents were obtained from participants and legal guardians/parents if the participant was under the age of 18. In privacy, subjects completed and sealed the questionnaire packet, and deposited it in a locked box at the health care office labeled "Q." The investigator retrieved the questionnaire packets at the end of the day. The investigator's contact information and a list of counseling and partner abuse services were attached to all questionnaire packets. Two individuals refused to participate in the study.

Measures

Decisions to select instruments for construct validity hypothesis testing were guided by scales measuring problems in interpersonal relationships and psychosocial well-being of abuse victims. Instruments selected were the Index of Peer Relations (IPR) (60- 61), the Miller Hope Scale (MHS) (62), and the Center of Epidemiologic Studies-Depression Scale (CES-D Scale) (63).

Burgess-Partner Abuse Scale for Teens (B-PAST). The 22-item scale is a self-report norm-referenced index, designed to measure the frequency of partner abuse in two dimensions: physical/ sexual (12 items) and social/emotional (10 items). The physical/ sexual partner abuse subscale is defined as a partner imposing a physical threat or actual bodily harm, with or without the use of a weapon, or a verbal or actual sexual behavior that threatens the teen girl's reproductive or sexual integrity. The social/emotional partner abuse subscale is defined as a partner statement or gesture that puts down, shames, humiliates, dictates, or demands what the teen girl can and cannot do regarding social activities. Each item is preceded by the stem "My partner" followed by responses such as "forces me to use drugs", "follows me when I do things with my family", or "won't let me go out my friends." Item response options are scored on a 5-point Likert type scale (range 0 to 4) and no items require a reversal in scoring. Interpretations of response options are left to the respondent. Because the domains are considered to be different yet related dimensions, item scores within each domain are summed to create subscales. Computed scores range from 0 to 88 for the total scale, 0 to 48 for the physical/ sexual partner abuse subscale, and 0 to 40 for the social/emotional abuse subscale. Higher scores reflect greater frequency of partner abuse. Severity of abuse has not been established. Flesch-Kincaid reading level is indicated at grade 2.3 (52).

The Index of Peer Relations (IPR). The IPR is a well validated self-report measure of the degree or magnitude of problems, dysfunction, or violence a teen has with his/her friends (60-61). The Likert questionnaire's 25 items ask adolescents (over age 12) to rate the frequency with which they experienced rejection, poor treatment by peers, or not getting along with peers. On a total score of 130, with item ranges from 0 to 6, higher scores indicate interpersonal violence and peer harassment. Flesch-Kincaid reading level is estimated at grade 3 (60-61, 64).

The Miller Hope Scale (MHS). The MHS is a widely used, reliable, and extensively validated self-report measure of the degree of hope and is defined by the anticipation for a future that is good and incorporates positive interpersonal relationships with others (62). The MHS consists of 40 items on a six-point Likert type scale (range 0 to 6) yielding a total score of 240. Higher scores indicate higher hope levels. However, no set score has been established to determine where hope is present or absent. Flesch-Kincaid reading level is estimated at grade 3 (62).

The Center of Epidemiologic Depression Scale (CESD). The CES-D Scale is a reliable and valid unidimensional selfreport measure consisting of 20 items designed to measure current (past 7 days) mood. A sad mood is described as the blues, loneliness, crying, sadness, or feelings of not being close or liked by others (63). The Likert type scale was developed out of a pool of several depression scales and interviews in the normal healthy population above age 12. Item responses range from 0 to 3 with higher scores indicating greater feelings of depression in the individual. No set score is established, though, for clinical depression. Flesch-Kincaid reading level is estimated at grade 2.1 (63).

Data Analysis Plan

Because the B-PAST (52) is a relatively new instrument, the internal consistency was examined by setting a criterion for the alpha coefficient above .70 (65-66). Poorly defined items were classified as those items that, when deleted, increased the alpha by more than .10. Items with correlations of less than .30 with the total scale score or items that negatively correlated with other items in the total scale or subscale were examined for their untoward effect on internal consistency (65-66).

Construct validity was examined by identifying and summarizing the major underlying dimensions of the B-PAST using exploratory factor analysis and standard procedures for identifying the factor structure (communality estimates, factor extraction, and promax rotation for the standard regression coefficients) (67). The factorloading cut off criterion was established at .40 and complex situations were defined when items loaded onto two factors showed differences in loadings of .2 or less (67). Promax rotation was selected in order to maximize the variability of items. The determination of meaningful factors was based on four criteria: (1) eigenvalues greater than one, (2) notable breaks in the scree test, (3) proportion of explained variance above 25% for each factor, and (4) conceptual interpretability (67). Further B-PAST construct validity was evaluated using the following criteria for correlations: p values less than .05 and Pearson correlations (68) of .30 or greater with the IPR scale (60-61) and the CES-D scale, (63) and less than -.30 with the MHS (62). One way ANOVA and differences in contrasted group means (abused vs. non-abused) were used to determine statistical differences (p values were less than .05) for interpersonal violence, depression, and hope (62).

In order to satisfy assumptions for factor analysis, contrasting group means, one-way ANOVA testing and Pearson correlations, the data were assessed using the following criteria: (1) Likert instruments were considered interval measurements; (2) participants were conveniently sampled but were obtained crossing ethnic, income, and geographical boundaries to achieve randomness and each score represented one observed variable; (3) the normal distribution was reviewed using frequency distributions. Data outliers were removed leaving a range of scores and standard deviations as low to moderate in size on the total instruments and their subscales; and (4) analyses remained robust because of the sample size of 239 (67-68).

Results

Validity

Responses from the teen sample were subjected to exploratory factor analysis using squared multiple correlations as prior communality estimates (67). In interpreting the rotated factor pattern, two factors emerged into the physical/sexual and the social/emotional dimensions (Table 2). Furthermore, dimensions were related yet distinct from the each other. Approximately 70.9% of the variance was explained by these two factors. Beyond factor two, the variance explained was insignificant.

Factor one, physical/sexual abuse, incorporates items related to actual or threats of physical and sexual partner abuse (i.e. "My partner kicks me, My partner says he will hurt me using a weapon, My partner gives me sex infections"). This factor explained the 37.5% of the variance and had the highest factor loadings. Its dominance is consistent with the literature and interactions involving a partner's threats or actual physical/sexual abuse. Factor two, social/emotional abuse, explained 33.4% of the variance, and items within this scale reflect behaviors consistent with a partner's tactics to humiliate, put down, demand, or dictate what another person can do (i.e. " My partner calls me bad names like bitch, My partner doesn't let me go out with my friends, My partner tells me what to wear").

As expected, findings demonstrated that the B-PAST had a negative correlation with the MHS (r = -.42, p < .001) and a positive correlation with the IPR (r = .39, p < .001) and the CESD (r = .46, p < .001). Statistical differences in contrasted group means (abused vs. non-abused) demonstrated that teen girls who were abused by a partner reported higher scores for abuse overall and for each subscale (n = 239, p < .0001). One-way ANOVA demonstrated that abused teens scored higher for violence in interpersonal peer relationships (n = 239, t = -2.50, p < .01) and more depression (n = 239, t = -4.09, p < .0001) but less hope (n = 239, t = 4.04, p < .0001).

Internal Consistency

The internal consistency coefficients for the 22-item B-PAST were .97, .95, and .91 for the total scale, physical/sexual abuse subscale, and the social/emotional abuse subscale, respectively. Item-to-total correlations for the 22-item scale ranged from .38 to .88. No items were found to increase the alpha by more than .10 if deleted. Additionally, no items were found to negatively correlate with items in the total scale or within its own subscale. The interfactor (physical/sexual subscale with the social/emotional subscale) correlation analysis revealed a moderate to strong correlation (r = .64), which suggested that each factor represented a related but distinct dimension of partner abuse. Table 4 outlines the reliability analyses.

Discussion

Compatible with the literature, the study confirmed that teen girls experience partner abuse (1-30, 34-39). The results also showed that the B-PAST extends the number of social and sexual acts perpetrated reported in the literature, which gives a more complete picture of teen partner abuse (1-26, 30, 35, 37).

Empirical support for hypothesis testing was provided by factor analysis, Pearson correlation, and one-way ANOVA or contrasted means to determine the instrument's ability to discriminate between teen girls who reported partner abuse and those who did not. A two-factor solution accounted for a high proportion of explained variance (70.9%) for the 22-item B-PAST. Other partner abuse scales have produced adequate explained variance but are limited to development and psychometric testing among adults (44-48). Pearson correlation showed a relationship between abused teen girls and depression, less hope, and peer interpersonal violence. These findings are consistent with the literature (2-26, 29-32, 34- 36, 49, 52). One-way ANOVA and differences between contrasted groups (abused vs. non-abused) were also consistent with the literature that showed teen girls abused by partners report higher scores for partner abuse, problems in other relationships (2-26, 29, 36, 49, 52), depression (30, 32), and less hope (31, 34).

Additionally, the findings lend support to the notion that social/ emotional abuse is an aspect of partner abuse that may be especially relevant to teens. In teen girls, social/emotional partner abuse is important to address since partners often do not live together, and therefore, physical abuse is more difficult to impose or keep hidden from the teen girl's family. The literature suggests that teen girls are more apt to comply with social control in an effort to please the partner (51-52).

Finally, the findings indicated that the B-PAST is well suited to measure partner abuse in teen girls. The Cronbach alpha coefficient was high suggesting a strong internal consistency. Based on a reading level of 2.3, the instrument was easy to read and complete, usually taking about five minutes.

Efforts were made to reduce the effect of non-probability sampling by obtaining adequate numbers of subjects across three settings to ensure representation of ethnic, income, and geographical groups. None-the-less, the subject to item ratio was well above the recommended number to conduct factor analysis.

In summary, the major contribution of this study was the psychometric assessment of a scale that can measure teen partner abuse, paying particular attention to social and sexual teen partner abuse. Because the B-PAST is designed to detect partner abuse in teen girls, it is anticipated that the recognition of partner abuse could aid clinicians to intervene or provide assistance to teen girls in dealing with this problem. Given that the results are limited to the scope of this study, further investigation is planned with similar and diverse samples of teens to assess replicability of these findings. A Hispanic translation as well as a back-translation of the B-PAST is in development to extend cultural relevance. Upon completion of these endeavors, the information gathered through the B-PAST can be used to identify risk factors and develop interventions.

Burgess-Partner Abuse Scale for Teens

Directions: During the past 12 months, you and one of your partners may have had a fight. Below is a list of things one of your partners may have done to you. Please circle the number of how often this partner did these things to you. This is not a test and there are no right or wrong answers. Remember, having a partner (s) does not mean you are having sex with the partner (s).

If you have not had a partner in the past 12 months, do not fill this form out.

Copyright, 2002 & 2003 Stephanie E. Burgess

Permission is granted for use in research or clinical settings. Burgess, S., Mackey, M., Felton, G., Tavakoli, A., & De Santis, E. Burgess- Partner Abuse Scale for Teens (B-PAST): Development and Initial Psychometric Assessment. Sigma Theta Tau Alpha Xi Viewpoints Research Conference, Columbia, SC; 2002.

REFERENCES

1. Coker AL, Smith P, McKeown R, King M. Frequency and correlates of intimate partner violence by type: Physical, sexual, and psychological Battering. American Journal of Public Health 2000;90:553-559.        [ Links ]

2. Levine S. The perils of young romance. U.S News & World Report 2001; 131: 46.        [ Links ]

3. Forshee V, Linder G, MacDougall J. Gender differences in the longitudinal predictors of adolescent dating violence. Preventive Medicine 2001;32:128-141.        [ Links ]

4. Forshee V, Linder G, Bauman K, Langwick S, Arriaga X, Heath J, McMahon P, Bangdiwala S. The safe dates project: Theoretical basis, evaluation design, and selected baseline findings. Youth Violence Prevention: Description and baseline data from 13 evaluation projects. American Journal of Medicine. 1996;12:39-47.        [ Links ]

5. Gamache D. Domination & control: The social context of dating violence. In: Levy, B., ed. Dating Violence: Young women in danger. Seattle: The Seal Press; 1991.        [ Links ]

6. Attala JM, Oetker D, McSweeney M. Partner abuse and female nursing students. Journal of Psychosocial Nursing 1995;33:17-24.        [ Links ]

7. Levy B. Dating violence: Young women in danger. Seattle: The Seal Press; 1991.        [ Links ]

8. Rickert V, Vaughan R, Wiemann C. Violence against young women: Implications for clinicians. Contemporary OB/GYN 2003;48:30-45.        [ Links ]

9. Suarez K. Teen dating violence: The need for expanded awareness and legislation. California Law Review 1994;82:423-471.        [ Links ]

10. James W, West C, Deters K, Armijio E. Youth dating violence. Adolescence 2000;35:455-465.        [ Links ]

11. O’Keefe M, Treister L. Victims of dating violence among high school students. Are the predictors different for males and females? Violence Against Women. 1998;4:195-223.        [ Links ]

12. Rickert V, Wiemann C,Berenson A. Health risks behaviors among adolescents with older partners. Archives Pediatrics & Adolescent Medicine 1997;151:276-280.        [ Links ]

13. Roberts T, Klein J. Intimate partner abuse and high-risk behavior adolescents. Archives of Pediatric & Adolescent Medicine 2003;157:375-380.        [ Links ]

14. Harrykissoon S, Rickert V, Wiemann C. Prevalence and patterns of intimate partner violence among adolescent mothers during the postpartum period. Archives Pediatrics & Adolescent Medicine 2002;156:325-330.        [ Links ]

15. Wiemann C, Agurcia C, Berenson A, Volk R, Rickert V. Pregnant adolescents: Experiences and behaviors associated with physical assault by an intimate partner. Maternal & Child Health 2000;4:93-101.        [ Links ]

16. Curry M, Doyle B, Gilholley J. Abuse among pregnant adolescents: Differences by development age. Maternal Child Nursing 1998;23:144-150.        [ Links ]

17. Dye T, Tolliver N, Lee R, Kenney C. Violence, pregnancy, and birth outcomes. Pediatric and Perinatal Epidemiology 1995;9:35-47.        [ Links ]

18. Jacoby M, Gorenflo D, Black E, Wunderlich C, Eyler A. Rapid repeat pregnancy and experiences of interpersonal violence among low-income adolescents. American Journal of Preventive Medicine 1999;16:318-321.        [ Links ]

19. Gessaner B. Experience of violence among teenage mothers in Alaska. JAMA 1998;280:492-501.        [ Links ]

20. Parker B. Abuse of adolescents: What can we learn from pregnant teenagers? AWHONNS Clinical Issues Perinatal Women’s Health Nursing 1993;4:363-370.        [ Links ]

21. Renker P. Physical abuse, social support, self-care, and pregnancy outcomes of older adolescents. Journal of Obstetrics and Gynecological Neonatal Nursing 1998;28:377-388.        [ Links ]

22. Rhynard J, Krebs M, Glover J. Sexual assault in dating relationships. Journal of School Health 1997;67:89-95.        [ Links ]

23. Wood K, Maforah F, Jewkes R. He forced me to have sex: Putting violence on adolescent sexual health agendas. Social Science Medicine 1998;47:233-242.        [ Links ]

24. Strauss S, Epeseland P. Sexual harassment and teens. Minneapolis: Free Spirit Publishing, Inc; 1992.        [ Links ]

25. Perrin K, Dindial K, Eaton D, Harrison V, Matthews T, Henry T. Responses of seventh grade students to “Do you have a partner with whom you would like to have a baby?” Psychology Reports 2000;86:109-118.        [ Links ]

26. Greenfield L, Rand M, Cvanen D. Violence by intimates: Analysis of data in crimes by current or former boyfriends, spouses, or girlfriends. Bureau of Justice Statistics Fact Book. Washington, D.C.: U. S. Department of Justice; 1998.        [ Links ]

27. Institute of Medicine. Leading health indicators for Healthy People 2010. Washington, DC: National Academy Press; 1999.        [ Links ]

28. Bureau of Justice Statistics. Trends in Juvenile Violence and the 1997 Update, publication number 170379 & 170377. Washington, D. C.: U. S. Department of Justice; 1997.        [ Links ]

29. Boris N, Heller S, Sheperd T, Zeanah C. Partner violence among homeless young adults: Measurement issues and associations. Journal of Adolescent Health 2002;30:355-363.        [ Links ]

30. Ackard D, Neumark-Sztainer D. Date violence and date rape among adolescents with disordered eating behaviors and psychological health. Child Abuse & Neglect 2002;26:455-473.        [ Links ]

31. Kashani J, Suarez l, Allan W, Reid J. Hopelessness in inpatient youths: A closer look at behavior, emotional expression, and social support. Journal American Academy of Child & Adolescent Psychiatry 1997;36:1625-1631.        [ Links ]

32. Kaslow N, Thompson M, Meadows L, Chance S, Puett R, Hollins L, Je S, Kellermann A. Risk factors for suicide attempt among African American women. Depression and Anxiety 2000;12:13-20.        [ Links ]

33. Kaslow N, Thompson M, Okun A, Price A, Young S, Bender M, Wy S, Twomey H, Goldin J, Parker R. Risk and protective factors for suicidal behavior in abused African-American women. Journal of Consulting Clinical Psychologist 2002;70:311-319.        [ Links ]

34. Bolland J, McCallum D, Lian B, Bailey C, Rowan P. Hopelessness and violence among inner-city youths. Journal of Maternal Child Health 2002;5:237-244.        [ Links ]

35. Bennett L, Fineran S. Sexual and severe physical violence among high school students. Power, beliefs, gender, and relationship. American Journal of Orthopsychiatry 1998;68:645-652.        [ Links ]

36. Smith H, Thomas S. Violent and nonviolent girls: Contrasting perceptions of anger experiences, school, and relationships. Issues in Mental Health Nursing 2000;21:547-575.        [ Links ]

37. Abbey A, Zawacki T, Buck P. Alcohol and sexual assault. Alcohol Research and Health 2001;25:43-51.        [ Links ]

38. Grisso J, Schwarz D, Hirschinger N, Sammel M, Brensinger C, Santanna J, Lowe R, Anderson E, Shaw, Bethel, C, Teeple L. Violent injuries among young women in urban area. New England Journal of Medicine 1999;341:1899-1905.        [ Links ]

39. Hoffmann J, Cerbone F, Su S. A growth curve analysis of stress and adolescent drug use. Substance Use and Misuse 2000;35:687-716.        [ Links ]

40. Durant T, Colley G, Saltzman L, Johnson C. Opportunities for intervention: Discussing physical abuse during prenatal care visits. American Journal of Preventive Medicine 2000;19:238-244.        [ Links ]

41. Clark K, Martin S, Petersen R, Cloutier S, Covington D, Buescher P, Beck-Warden M. Who gets screened during pregnancy for partner violence? Archives of Family Medicine 2000;9:1093-1099.        [ Links ]

42. Nelson H, Nygren P, McInerney Y, Klein J. Screening for women and elderly adults for family and intimate partner violence: A review of the evidence for the U.S. Preventive Services Task Force. Annals of Internal Medicine 2004;140:387-396.        [ Links ]

43. Phillips D. Culture and systems of oppression in abused women’s lives. Journal of Gynecological & Neonatal Nurses 1998;27:678-683.        [ Links ]

44. Campbell J. Nursing assessment for risk of homicide with battered women. Advances in Nursing Science. 1986;4:36-51. 45. Hudson W. The Partner Abuse Scale. Tempe, Arizona: Walmyr Publishing; 1992.        [ Links ]         [ Links ]

46. Hudson W, McIntosh S. The assessment of spouse abuse: Two quantifiable dimensions. Journal of Marriage and the Family 1981;43:873-888.        [ Links ]

47. Straus M. Measuring intrafamily conflict and violence: The Conflict Tactics Scales. Journal of Marriage and the Family 1979;41:75-88.        [ Links ]

48. Marshall L. Development of the Severity of Violence Against Women Scales. Journal of Family Violence 1992;103-121.        [ Links ]

49. Parker B, McFarlane J, Soeken K, Torres S, Campbell D. Physical and emotional abuse in pregnancy: A comparison of adult and teenage women. Nursing Research 1993;42:173-354.        [ Links ]

50. Parker B, McFarlane J, Soeken K. Abuse during pregnancy: Effects on maternal complications and birth weight in adult and teenage women. Journal of Obstetrics & Gynecology 1994;84:323-328.        [ Links ]

51. Walker LE. The battered woman syndrome. New York: Springer Publishing Company; 2000.        [ Links ]

52. Burgess S, Mackey M, Felton G, Tavakoli A, De Santis E. Burgess-Partner Abuse Scale for Teens (B-PAST): Development and Initial Psychometric Assessment. Sigma Theta Tau Alpha Xi Viewpoints Research Conference, Columbia, SC; 2002.        [ Links ]

53. Orem D. Concepts of Protection (3rd ed.). New York: McGraw- Hill Publishers; 1985.        [ Links ]

54. Cunningham A, Jaaffe P, Baker L, Dick T, Malka S, Mazaberi N, Poisson S. Theory-drives explanations of male violence against female partners. Literature update and related implications for treatment and evaluation. [On-line] www.lfcc.on.ca/maleviolence.pdf. Retrieved January 2005.        [ Links ]

55. Straus M, Gelles R. Physical violence in American families. New Brunswick: Transaction Publishers; 1990.        [ Links ]

56. Heise L. Violence against women: An integrated, ecological framework. Violence Against Women 1998;4:262-290.        [ Links ]

57. World Health Organization. World report on violence and health. Geneva, Switzerland: World Health Organization; 2002.        [ Links ]

58. Arias I, Paper K. Psychological abuse: Implications for adjustment and commitment to leave violent partners. Violence and Victims 1999;14:555-567.        [ Links ]

59. Campbell J. Beating of wives: A cross cultural perspective. Victimology 1985;10:174-185.        [ Links ]

60. Nurius P, Hudson W, Daley J, Newsome R. A short form scale to measure peer relations dysfunction. Journal of Social Science Research 1988;12:27-29.        [ Links ]

61. Klein W, Beltran M, Sowers-Hoag K. Validating an assessment of peer relationship problems. Journal of Social Science Research. 1990;13:71-85.        [ Links ]

62. Miller J, Powers M. Development of an instrument to measure hope. Nursing Research 1988;37:6-10.        [ Links ]

63. Radloff L. The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement 1977;1:385-401.        [ Links ]

64. Hudson W. The Walmyr assessment scale scoring manual. Tallahassee, Florida: Walmyr Publishing Company; 1997. 65. Nunnally J, Bernstein I. Psychometric theory. 3rd ed. New York: McGraw-Hill Company; 1994.        [ Links ]         [ Links ]

66. Carmines E, Zeller R. Reliability and Validity Assessment. Newbury Park: Sage Publications; 1979.        [ Links ]

67. Hatcher L. A step-by-step approach to using the SAS system for factor analysis and structural equation modeling. Cary, N.C.: SAS Institute; 1994.        [ Links ]

68. Hatcher L, Stepanski E. A step-by-step approach to using the SAS system for univariate and multivariate statistics. Cary, N.C.: SAS Institute; 1994.        [ Links ]