Parent ratings of depression, anxiety, and aggression in elementary school and impatient 8to 12-year-olds were evaluated by examining their correspondence and discrepancies with parallel child and teacher ratings. Parent-child and parentteacher correspondence was significantly better for elementary school than inpatient children on depression and anxiety. Parents and teachers reported similar levels of symptoms on all traits for elementary school children, yet parents reported significantly more symptoms than teachers in the inpatient sample. The level or severity of anxiety symptoms that parents reported in inpatient children was similar to that reported by children, yet inpatient children reported significantly Iess depression and aggression than parents. Elementary school children reported significantly more symptoms on all traits than parents. Age, sex, and sample differences in discrepancies were examined. 0 1996 John Wiley & Sons, Inc.
A well known finding in the assessment of child psychopathology is the low correspondence that is exhibited between reports gathered from diverse informants. In particular, a meta-analysis found parent and child reports are only moderately related ( r = .25), as are parent and teacher reports ( r = .27; Achenbach, McConaughy, & Howell, 1987). Efforts at understanding this low correspondence and identifying factors that may contribute to it have only recently received more attention (cf., Kolko & Kazdin, 1993).
The relationship between informant's reports may differ according to the type of behaviors assessed, the sample employed, and the measures used in the assessment. In addition, agreement may vary across age groups and sexes. Furthermore, the severity or number of problems reported may also differ between informant's reports, and this variability may be a function of the above features. Although the meta-analysis and extensive work on the Achenbach parent (CBCL; Achenbach, 1991a), teacher (TRF; Achenbach, 1991b), and youth self-report (YSR; Achenbach, 1991c) scales have provided important data on the above factors both across studies and with reference to the Achenbach scales, few investigations have addressed many of these factors within a single study. Furthermore, the data from the YSR scales is confined to clinic-referred and nonreferred 11 to 18-year-olds, and may not be generalizable to inpatient samples and younger children.
Others have found better agreement across informants on scales that are parallel or equivalent in content across the informants (Epkins, 1993;Epkins & Meyers, 1994). Also, Lachar recently developed a self-report scale that parallels dimensions derived from the Personality Inventory for Children (PIC; Lachar, 1982;Wirt, Lachar, Klinedinst, & Seat,