Abstract
The construct and concurrent validity of the Roberts Apperception Test for Children (RATC) was investigated for children of two ethnic groups and for males and females. The sample included 294 children ages 6-15 drawn from two school districts located in Texas and Utah. The results did not provide conclusive evidence for the similarity of factor structures of the RATC across the two ethnic groups and between sexes. The underlying theoretical constructs of the RATC do not appear to be equivalent for Anglos and Hispanics, females and males. The RATC does appear to be both a measure of adaptive functioning and clinical indicators and is influenced by developmental trends. However, for separate groups individual subscales contribute differentially to the underlying constructs being assessed. Investigations of bias in the concurrent validity of the RATC revealed that, with one exception, there was no constant or systematic error in the estimation of Child Behavior Checklist (CBCL) scores as a function of group membership. Evidence was provided for bias in the prediction of the Internalizing score, Teacher's form, as a function of ethnic group membership. Teachers rated Hispanic students higher on the Internalizing scale than Anglo students. Cautious use and interpretation of the RATC is recommended until further research is conducted. Use of the RATC should be tempered with the knowledge that for different groups individual subscales contribute differentially to the understanding of a child's adaptive functioning.
Burns, Christine Wasiluk (1986). The validity of the Roberts Apperception Test for children across ethnic groups and between sexes. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -609126.