Adolescent Clinical Sexual Behavior Inventory-Self Report
Friedrich, W.N., Lysne, M., Sim, L., & Shamos, S. (2004). Assessing sexual behavior in high-risk adolescents with the Adolescent Clinical Sexual Behavior Inventory. Child Maltreatment, 9(3), 239-250.
The ACSBI is a screening measure designed for clinical populations to assess sex-related behaviors that might suggest a need for intervention. It assesses sexual risk taking, nonconforming sexual behaviors, sexual interest, and sexual avoidance/discomfort.
The ACSBI is based on the Child Sexual Behavior Inventory (CSBI), a widely used measure of child sexual behavior for
children aged 2-10, which is also reviewed in this database. Although the measure is psychometrically young and in need of some revision, it is an important measure in that risky and unsafe sexual behaviors are important to assess in adolescents, especially those with histories of sexual abuse.
There are two versions of the ACSBI, a parent-report version (ACSBI-P) and an adolescent self-report version (ACSBI-S). Both are reviewed in this database.
3-point scale (1=not true, 2=somewhat true, 3=very true)
|Sexual knowledge/interest||Shows off their skin or body parts.|
|Sexual risk/misuse||You are worried about your sexual behavior.|
|Divergent sexual interests||Peeps into windows or ties to see others in the bathroom.|
|Concerns about appearance||Is unhappy with their looks.|
|Fear||Has no friends of the opposite sex.|
1. The ACSBI is based on the Child Sexual Behavior Inventory, which assesses sexual behaviors in children aged 2-10 and is also reviewed in this database.
2. The ACSBI-S is an adolescent self-report version for adolescents (also reviewed in this database).
Friedrich et al. (2004) reported on the initial reliability and validity of the measure. Some scales showed questionnable internal consistency.
|Test-Retest- # days: 7||Acceptable||r||0.74||0.74||0.74|
From Friedrich et al. (2004):
Conducted with 23 inpatients was reported for the ASCBI-S total score.
Reliability for scales varied and was reported as:
Sexual Knowledge/Interest (alpha=.84)
Sexual Risk/Misuse (alpha=.77)
Concerns About Appearance (alpha=.68)
Divergent Sexual Interest (alpha=.65)
Reliability for the total scale was .86, with all items positively correlated with the total score.
Correlations between the parent and adolescent reports using the ACSBI-P and ACSBI-S
showed that 28 of 41 items were correlated. Total scores were correlated (r=.50).
However, the authors note that the ACSBI-P and ACSBI-S share only 25% of their
variance, suggesting that parents and adolescents have different perceptions regarding
adolescents’ sexual behaviors.
From Friedrich et al. (2004):
The ACSBI is based on the Child Sexual Behavior Inventory (CSBI), a widely used measure of child sexual behavior for children aged 2-10, which is also reviewed in this database. It is based on 2 domains of the CSBI, which are elevated for 10- to 12-year-olds.
The initial pool of 35 items assessed sexual knowledge and interest. Other items were added “to reflect salient behaviors described in the literature, for example, sexual concerns, promiscuity, body image, sexual risk taking, and running away.”
Items were tested with 23 inpatient adolescents. “Troublesome items were reworded, and 10 items were added.”
|Validity Type||Not known||Not found||Nonclincal Samples||Clinical Samples||Diverse Samples|
|Sensitive to Change|
|Sensitive to Theoretically Distinct Groups||Yes|
From Friedrich et al. (2004):
ACSBI-S scores correlated with the three Sexual Concerns scales of the Trauma Symptom Checklist for Children (TSCC), the Adolescent Sexual Concerns Questionnaire Sexual Distress Scale (ASCQ), and the CBCL total and subscale scores.
Adolescents with a sexual abuse history had higher scores on the ASCBI-S total score and Sexual Knowledge/Interest, Sexual Risk/Misuse, and Fear/Discomfort scales compared to those without a history of abuse. Friedrich et al. (2004) examined the factor structure of the ACSBI using a
principal components analysis and an orthogonal varimax rotation. They identified 5 factors for both the parent and self-report version. For the ACSBI-S (self-report version), the factors accounted for 37.6% of the variance. Of the 45 items, 39 were included in the solution, as they had item loadings greater or equal to .30 on at least one scale.
Factors were labeled: 1) Sexual Knowledge/Interest, 2) Sexual Risk/Misuse, 3)Divergent Sexual Interest, 4) Concerns About Appearance, and 5)Fear/Discomfort. These are the same factors identified for the parent-report version (ACSBI-P).
From Wherry et al. (2009):
The primary goal of this study was to determine if the ACSBI conformed to the five-factor scale format that was initially used with the clinical sample (including those referred for sexual abuse evaluations. The principal components analyses found a different though similar set of factors. More specifically, the ACSBI scales reflected three of the five scales with similar items: Concerns about Appearance, Sexual Knowledge, and Sexual Deviance. The two factors that did not emerge included Sexual Risk-Taking and Fear/Discomfort. As noted, while the sample was similar, it did not include a sample of adolescents presenting for sexual abuse evaluations. While the number of items and factors was reduced, the three scales accounted for a greater percentage of the total variance (41.6%) than did the original Friedrich et. al. (2004) study (37.6%). Furthermore, these ACSBI-S factors demonstrated adequate reliability with alpha coefficients ranging from .61 to .75.
|Not Known||Not Found||Nonclinical Samples||Clinical Samples||Diverse Samples|
No information available.
1. As noted by the authors, the Fear/Discomfort scale has questionable reliability.
2. Only 29 of 45 items loaded on scales in the factor analysis.
3. Psychometrics examined in a predominantly White, middle- to upper-class sample. More research is needed on psychometrics, but this is a promising measure.
4. The author reports that the measure is being used in two longitudinal studies at two different sites: 1) Medical University of South Carolina (Elizabeth Letourneau), and 2) University of North Carolina at Chapel Hill (LONGSCAN).
The development sample (Friedrich et al., 2004) included 174 adolescents aged 12-18 (M=15, SD=1.4) and their primarily female caregivers (81.5%). The sample was predominantly White, educated (74% had 12+ years of education), and upper- to middleclass. Adolescents included 46.6% males and 53.4% females. Adolescents were recruited from a partial hospitalization or inpatient program at the Mayo Clinic in Rochester, New York(n=120). or from an outpatient evaluation program in Denver, Colorado (n=54).