Teacher behavior ratings of adolescents in a residential treatment facility: The role of client gender and ethnicity
Steck Silvestri, Erin L
MetadataShow full item record
Behavior rating scales have demonstrated their utility in their inclusion of a wide variety of symptoms and behaviors, ability to allow for normative comparison, and time/cost efficiency. Despite these benefits, behavior rating scales have been criticized for their frequent use of arbitrary cut off scores, inconsistent definition of constructs, and failure to adequately define language used ( often, sometimes, never ). These limitations likely carry unique implications for ethnic minority groups, based on the recurrent finding that individual characteristics of the child such as ethnicity or gender might unduly influence the ratings of adult informants. Teacher ratings on a sample of 65 adolescents in a residential treatment facility were gathered with the Conners' Teacher Rating Scale, Revised, Long Form (CTRS-R:L). Independent samples t -tests revealed no significant mean differences between internalizing (INT) and externalizing (EXT) diagnostic groups on CTRS-R:L measures of ADHD, ODD, or MOOD. Chi square analyses indicated that minority (MIN) adolescents were significantly more likely to be diagnosed with EXT diagnoses than their majority (MAJ) counterparts. In addition, males were more likely to receive EXT diagnoses than their female peers. A logistic regression revealed that although teacher ratings of adolescents' behavior did not contribute significantly to predictive power with respect to primary diagnostic group membership, there was a significant interaction effect for ethnic status and gender. In particular, the odds of belonging to the INT diagnostic group being significantly higher for MAJ females than for any other ethnic/gender group. The findings are discussed in relation to the larger context of behavioral assessment with ethnically diverse children and adolescents.