Submitted by mholliday on Thu, 07/12/2012 - 11:53
Bell, M.D. (2005) Bell Relationship Inventory for Adolescents (BRIA) Manual. Los Angeles: Western Psychological Services.
The BRIA is an adapted version of the Bell Object Relations and Reality Testing Inventory (BORRTI) that was developed for adolescents aged 11 to 17 and assesses for psychological disturbances and relationship problems. The BORRTI has been widely used with adults and is reviewed in this database. The BRIA contains only the Object Relations subscales (and not the Reality Testing subscales) from the BORRTI.It yields scores on 4 scales comparable to the BORRTI scores: 1)Alienation, 2) Insecure attachment, 3) Egocentricity, and 4) Social incompetence. It also yields a score on one additional scale Positive Attachment, which reflects healthy object relations. While both measures have validity scales, the BRIA has a response bias scale in place of the three scales found on the BORRTI. The measure can be administered to individuals orgroups.Content from the BRIA copyright (c) 2005 by Western Psychological Services. Reprinted for reference within the NCTSN Measure Review Database by permission of the publisher, WPS, 12031 Wilshire Boulevard, Los Angeles, CA 90025, www.wpspublish.com. All rights reserved.
The Bell Object Relations and Reality Testing Inventory (BORRTI) is the original version on which the BRIA is based. It contains reality-testing subscales in addition to the object relation scales found on the BRIA. Object relation scale items are comparable except that for the BRIA, three items were dropped and others were added. The wording on most items was simplified on the BRIA.
The 625 public school students who were part of the development sample (see “Population Used to Develop Measure”) comprised the referencesample used to develop standard scores. Only those with no missing items were used. This group averaged 14.3 years of age (SD=1.85, Range=11-19); 273 were boys (44%) and 348 were girls (56%). Ethnicity was 65% White, 19% Black/African American, 7% Hispanic, 3% Asian, 4% Other, and2% Missing.One set of norms exists for boys and girls across all age groups and ethnicities. The author examined differences and found some significantdifferences for gender and age but reasoned that the effect sizes are small. One ethnicity difference in the moderate-effect size range was found forHispanic youth compared to White youth, suggesting that further research may be needed with this population to determine whether the differencerepresents a true population difference.
From Bell (2005)For the development sample, internal consistencies were .76 for Positive attachment, .69 for alienation, .76 for insecure attachment, .69 for Egocentricity, and .77 for Social incompetence.There were 59 community college students, and they completed both the BRIA and BORRTI. Correlations for comparable scales were as follows: Alienation = .59, p<.0001; Insecure Attachment = .75, p<.0001; Egocentricity = .43, p<.0001; Social Incompetence = .83, p<.0001.As the author states, caution must be made when comparing an adolescent's score on the egocentricity scale on the BRIA to that same score on the BORRTI. The egocentricity and alienation scales on the two forms (BRIA and BORRTI) may not be parallel.
The BRIA was developed by adapting items from the BORRTI Form O. Focus groups were conducted with children in grades 6-9, leading to item revision. Items were changed to simplify language and because focus group feedback suggested that children were interpreting the items differently from how they were meant to be interpreted.Sixteen new items were added to capture relationship experiences that are psychologically meaningful to children, including aspects of their relationships with peers and parents.
From Bell (2005), Bell (2003)Concurrent validity examined by correlating BRIA scale scores and scores from the Personality Inventory for Youth and the Behavior Assessment System for Children: Self-Report for Adolescents.Examination of mean BRIA scores for different groups (residential psychiatric, residential family treatment, outpatient, and non-clinical) showed significant differences between these groups.A factor analysis was conducted using the development sample, which included clinical and non-clinical populations. Principal components analysis with an oblique rotation yielded 5 factors, accounting for 16% of the variance. Four of the factors were identical to those found for the adults’ responses on the BORRTI Form O. Analyses examining stability of factor structure found good congruence coefficients for all 5 factors.
1. While a factor analysis has been conducted, the variance accounted for by this factor solution is low (16%).2. Internal consistencies are acceptable, but this would be expected, given that they were calculated using the same sample with which the factor structure was determined. In addition, while acceptable, the internal consistencies are in the lower range of acceptable. Additional work is needed to examine whether the internal consistencies are similar whenexamined in other populations.3. While the development and normative samples did include a reasonable percentage of African-Americans (19%), it did not involve large samplings of other ethnic groups. More research may be needed before applying the measure to other groups, especially given the author's caution regarding using the BORRTI with respondents who are not fluent in English because it contains idiomatic expressions that are not well understood by foreign-bornEnglish speakers.
Sample: 815 children (59.9% girls and 46.1% boys), of which 705 were recruited from public schools and 110 came from clinics and residential treatment centers.
The ethnic composition was as follows: 64.5% White, 19.4% Black/African American,8.8% Hispanic, 3.1% Asian, and 4.2% Other. Average age was 14.4 (SD=1.9, Range 11-17). Average grade in school was 8th grade (SD=1.94, Range=6-11).
1. The measure is based on a widely used measure of adult relationship functioning. Although that measure could be used with adolescents, the author chose to refine the measure and to develop this version, with wording and norms that are more appropriate for this population.2. The measure taps potentially important therapeutic dimensions that may be especially relevant when working with children exposed to trauma.
1. The measure is relatively new and is not widely used.2. While a factor analysis has been conducted, the variance accounted for by this factor solution is low (16%).3. Internal consistencies are acceptable, but this would be expected, given that they were calculated using the same sample with which the factor structure was determined. In addition, while acceptable, the internal consistencies are in the lower range of acceptable. Additional work is needed to examine whether the internal consistencies are similar when examined in other populations.4. While the development and normative samples did include a reasonable percentage of African-Americans (19%), it did not involve large samplings of other ethnic groups. More research may be needed before applying the measure to other groups, especially given the author's caution regarding using the BORRTI with respondents who are not fluent in English because it contains idiomatic expressions that are not well understood by foreignbornEnglish speakers.
The reference for the manual is:Bell, M.D. (2005). Bell Relationship Inventory for Adolescent (BRIA) Manual. Western Psychological Services, Los Angeles, CA.A PsychInfo search (6/05) for "Bell Relationship Inventory for Adolescent Manual" or "BRIA" anywhere revealed that that the measure has been referenced in 1 peer-reviewed journal article.1. Bell, M.D. (2003). Bell Object Relations Inventory for adolescents and children: Reliability, validity, and factorial invariance. Journal of Personality Assessment, 80(1), 19-25.