Today's technologies, including videoconferencing software, video games, and virtual reality, have become integral to education. These technologies allow individuals to represent themselves through virtual identities. However, these virtual identities can inadvertently trigger stereotype threat--a psychological state where individuals feel at risk of confirming negative stereotypes related to their social group and can underperform educationally as a result. Thus, addressing stereotype threat in virtual environments is crucial. This project aims to create computational interventions and virtual representations that reduce stereotype threat in online educational settings, thereby contributing to the development of a diverse and globally competitive workforce. The specific objectives of this research within the context of educational settings are: 1) Investigate the impact of different visual and auditory virtual identities on stereotype threat, and 2) Develop avatar customization systems that mitigate stereotype threat by leveraging insights from the first objective. Through controlled studies, this project generates new insights into several areas: 1) The effects of different visual and auditory virtual identities on stereotype threat among diverse online learners, 2) How to develop avatar customization systems aimed at reducing stereotype threat in online educational settings, 3) The effectiveness of these systems in improving performance for diverse learners, and 4) Gen