BACKGROUND
This study explored the use of interface agents, anthropomorphic, 3D-animated computer characters  that provide  teaching or mentoring within a computer-based learning environment, to encourage young Black and White women to pursue careers in engineering.
PURPOSE (HYPOTHESIS)
We hypothesized  that computer-based models  that matched young women in  terms of  their  race  and gender would be  the most  effective  in positively influencing their interest, self-efficacy, and stereotypes about engineering.
DESIGN/METHOD
Eighty African American undergraduate female students in Experiment 1, and 39 White undergraduate  female  students  in Experiment 2  interacted with a computer-based agent  that provided a narrative designed  to encourage  them to pursue engineering careers. The study employed a 2 3 2 between subjects factorial design  (agent gender: male vs.  female  and  agent  race: Black vs. White).

RESULTS
Across both studies we found that race and gender influenced the effectiveness of the agent for several key outcome measures. Computer-based agents who matched  the participants with  respect  to  race  and gender  tended  to be  the most  effective  in  improving  the women’s  responses  to  engineering-related fields. Nevertheless,  the White male  agent was  actually  significantly more influential than the White female agent for female Black participants.
CONCLUSIONS
Personalizing  interface  agent  characteristics  to match  the  target population can  increase  the  effectiveness of  a persuasive message  to  encourage young women  to pursue  engineering. Such  an  approach may  contribute  to  the growth and inclusiveness of fields such as engineering.