Teacher bias in the virtual classroom

Abstract

The purpose of this preregistered study was to examine whether patterns of teachers' race and gender biases found in prior research would emerge in a virtual teaching context. We asked a national sample of 989 teachers to evaluate student work pictured in screenshots of a virtual classroom showing students explaining their work. We randomly assigned the images of students depicted in the virtual classroom and tested whether the race or gender of the student would trigger biases with regard to teachers' evaluations of correctness, student ability, and recommendations that students be tested for gifted programs or special education services. We found no evidence of correctness or ability bias, but that teachers recommended male students more often to be tested for gifted programs and Black students for IEPs based on the same student work. We also conducted exploratory analyses revealing that school enrollment rates of Black students moderated teachers' ability and tracking biases.