Abstract
An experiment with 146 Kindergarten, first-, fourth-, and fifth-grade children was conducted to determine whether the findings of previous mediation research can be generalized to a new television context (gender stereotyped television) and a new population (African American children). A 3 (mediation type) × 2 (grade) × 2 (race/ethnicity) experiment revealed that the results of prior work can be replicated among Caucasian children but not among African American children. Future work is encouraged to study non-Caucasian children to increase mediation research's theoretical scope and ability to address the needs of a larger, more diverse population.