
The researchers created fake accounts for students in 124 massive open online courses (MOOCs). The names associated with the accounts were designed to give a strong indication that students were either White, Black, Indian, or Chinese. These fictional students submitted online posts or questions to the instructors.
The researchers found that instructors in these MOOCs responded to about 7 percent of all inquiries posted. But for the fictional accounts designed to give the impression the student was a White male, the instructors responded 12 percent of the time.
The authors concluded that “our field experiment produced evidence that the comparative anonymity granted by asynchronous, digitally mediated interactions in online discussion forums does not eliminate bias among instructors. Indeed, we found a sizable bias in favor of White male identities which were nearly twice as likely to receive a discussion-forum response from the instructor compared to other student identities.”
The full report, “Bias in Online Classes: Evidence from a Field Experiment,” may be downloaded by clicking here.

