This project addressed two major goals in a sample of adolescent girls in the child welfare system in the Denver, Colorado metro area. First, researchers compared the effectiveness of two interventions for revictimization among adolescent girls in the child welfare system using a randomized, longitudinal design. Second, researchers examined process variables targeted by the respective interventions longitudinally.
Adolescents received a flyer about the Healthy Adolescent Relationship Project (HARP) via their
caseworker, foster parents, or service providers. For adolescents under age 18 interested in
participating in the project, parental or child welfare administrative consent was secured (depending on
the custody status of the young woman) prior to contacting the adolescents to invite them to the Time 1
assessment. Adolescents aged 18 and 19 either contacted the research team director or gave permission
for their contact information to be given to the research team. Upon receiving referrals, research staff
initiated phone calls to invite potential participants. During the initial phone call, adolescents were told
that the project involved four interviews as well as participation in a 12-week healthy relationship class.
If adolescents were interested, they were invited to attend the Time 1 (3-hour) assessment.
Participants were greeted by a graduate-level interviewer or the project manager. During the
assent/consent process, adolescents were informed in both written and verbal formats about the scope
of the study (including that they would be asked about exposure to interpersonal trauma). The
interviewer administered an "assent/consent quiz" designed to assess understanding of the
assent/consent information during the assent/consent process. Adolescents were considered
assented/consented into the study if they answered the quiz questions correctly and provided written
assent/consent (depending on their age). Following assent/consent procedures, participants completed
the interview. At the end of the session, participants were asked to complete questions to monitor
responses to study procedures.
Prevention groups were started on a rolling base, once approximately sixteen teens had
completed the Time 1 assessment. Each intervention session lasted approximately one and a half hours;
12 sessions were implemented weekly. Teens received $10 after each group session to help cover
transportation costs. To encourage attendance, teens were also entered into a raffle drawing once they
had attended 9 sessions. For each additional session after 9 sessions they received one additional raffle
entry. After the cohort completed the 12-week prevention group protocol, a drawing was conducted for
one winner of a $50 gift card or an iPod shuffle.
After the 12-week curriculum was implemented, participants were invited back for three
additional 2-hour assessments: immediately post-intervention group, 2- and 6-months after the
intervention group. Notably, adolescents who did not attend the intervention were also invited back to
the three follow-up interviews. The post-group assessment occurred as close to the last intervention
meeting as possible. Pre-, post-, 2-month, and 6-month assessments were all administered one-on-one
by graduate level research staff who were blind to randomization condition. At the end of each
interview, teens were compensated $40 for their time and $10 to help cover transportation costs. At the
end of the Time 1 interview, teens were offered a newsletter that provided referrals to community
agencies dealing with health and violence issues.
Adolescent females between the ages of 12 and 19 who were currently or previously involved with the child welfare system were referred by their case workers, service providers, or legal guardians based on their child welfare involvement and history of maltreatment exposure. The research team
worked with multiple counties as well as community agencies that serve child welfare youth in Denver
and surrounding areas to recruit participants into the study. Additional inclusion criteria were: 1) no
current suicidal ideation; 2) current treatment services if teens reported suicide attempts or psychiatric
hospitalizations in the last three to six months; and 3) current treatment services if teens reported
current self-harm behavior or psychosis. Based on these criteria, 214 referrals were received, of whom 180
completed an interview at Time 1. Of these 180, 4 did not meet the inclusion criteria (e.g., due to
current psychotic symptoms) and were excluded from the intervention and Time 2-4 assessments.
At the end of the pre-treatment assessment, all adolescents were randomly assigned to two groups: one social learning/feminist (SL/F) and one risk detection/executive function (EF/RD). Group start dates were set on a rolling basis as researchers completed enough pre-assessments to populate the groups. Fifteen adolescents never confirmed to attend a group; thus, these participants were never randomized.
Longitudinal: Cohort / Event-based
Adolescent females between the ages of 12 and 19 involved with the child welfare system in the Denver, Colorado, metro area.
The study contains one data file with 180 cases and 545 variables.
- Information on trauma exposure, dating violence, and revictimization was collected using the Traumatic Events Screening Inventory-Child version (TESI).
- Dating violence was assessed using the Conflict in Adolescent Dating Relationship Inventory (CADRI), a 70-item measure that assessed the frequency with which conflict tactics are used with dating partners.
- The presence or absence of revictimization at follow-up interviews (Time 2, 3, and 4) was calculated using a combination of data from the TESI and CADRI.
- Information on acceptability of dating violence was collected from an adapted vignette task created by Cauffman, Feldman, Jensen, and Arnett to assess teens' understanding of dating violence.
- Knowledge about gender roles and sexism were collected using The Ambivalent Sexism Inventory (ASI), a 22-item self-report measure that assessed two types of sexism: hostile and benevolent.
- A lexical decision making task was administered to assess relationship expectancies.
- To collect information regarding knowledge about potential responses to danger, teens read vignettes adapted from the Salt Lake City Teen Dating Violence Toolbox.
- Belief in capacity for social action was measured with The Sociopolitical Control Scale (SPCS) which measures the capacity for individuals to act in their communities.
- The Adolescent Interpersonal Competence Questionnaire (AICQ) was used to assess healthy relationship skills and attitudes across four competence domains: 1) dating relationships; 2) providing emotional support to friends; 3) management of conflicts; and 4) assertion.
- Researchers administered a modified version of the Assertion Inventory (AI) that assessed the likelihood of engaging in assertive behaviors across three domains: substance use, general, and social situations.
- The Communication Skills Test (CST) described both negative and positive communication patterns.
- Risk detection was assessed using a Wason Selection Task (WST) in which participants were asked to detect violations of social and safety rules in dating situations.
- Emotion awareness was assessed using the Difficulties in Emotion Regulation Scale (DERS), a 36-item measure designed to assess multiple aspects of emotion regulation, including: non-acceptance of emotional responses; difficulties engaging in goal directed behaviors; impulse control difficulties; lack of emotional awareness; limited access to emotion regulation strategies; lack of emotional clarity.
- Several domains of executive function were assessed using behavioral measures, such as set shifting; interference control; and working memory.
- To assess for inattention symptoms participants completed 11 of the 18 items of the Adult ADHD Self-Report Scale (ASRS), a self-report measure based on DSM-IV criteria for Attention Deficit Hyperactivity Disorder.
- At each time point, researchers assessed the frequency and quantity of teens' drug and alcohol use in dating situations over the previous month.
- The Trauma Symptom Checklist for Children (TSCC) assessed symptoms commonly associated with the experience of traumatic events across various domains.
- Depression symptoms were assessed with the Beck Depression Inventory 2 (BDI-II).
- The Marlowe-Crowne Social Desirability Scale was administered at the second interview to control for social desirability tendencies.
- To assess adolescent alliance with clinicians, researchers administered the Therapeutic Alliance Scale for Adolescents (TASA), a 12-item self-report measure, at the end of Meetings 2, 6, 10, and 12.
- To assess group processes, researchers administered the Intervention Group Environment Scale (IGES) at Meetings 2, 6, 10, and 12. This 25-item measure comprised three subscales: 1) implementation and preparedness; 2) counterproductive activity; and 3) cohesion.
Based on sampling criteria, researchers received 214 referrals, of whom 180 completed an interview at Time 1 for a response rate of 84 percent.
This study used the following:
- Acceptability of dating violence (ADV)
- Adolescent Interpersonal Competence Questionnaire (AICQ)
- Adult ADHD Self-Report Scale (ASRS)
- Ambivalent Sexism Inventory (ASI)
- Assertion Inventory (AI)
- Beck Depression Inventory 2 (BDI-II)
- Communication Skills Test (CST)
- Conflict in Adolescent Dating Relationship Inventory (CADRI)
- Difficulties in Emotion Regulation Scale (DERS)
- Intervention Group Environment Scale (IGES)
- Marlowe-Crowne Social Desirability Scale
- Sentence Span Task (SST)
- Sociopolitical Control Scale (SPCS)
- Therapeutic Alliance Scale for Adolescents (TASA)
- Trauma Symptom Checklist for Children (TSCC)
- Traumatic Events Screening Inventory-Child version (TESI)
- Wason Selection Task (WST)