Online learning platforms make education- including courses on the latest technologies widely accessible to those who cannot afford prestigious academic institutions. How learners rate online instructors affects service delivery decisions on these platforms. Learning outcomes may also differ based on who delivers online content. Differences- or biases- in how learners rate instructors and whom they learn best from can hurt the diversity of both instructors and learners, and reduce learning effectiveness, resulting in inequitable access to the technologies of the future. The researcher’s observational data reveal gender bias towards female instructors in teaching ratings, particularly for tech courses, including AI and more broadly machine learning courses. Additionally, prior ratings significantly affect learners’ evaluation of the instructor’s performance across prior online experiments.

The researchers are collaborating with Coursera via LBS Ed Tech to run a field experiment building on prior findings, by designing and running a Coursera course that includes a technical module and a qualitative module. They will examine bias in online teaching, induced by prior ratings and instructor gender, in a paid Coursera course that includes both a quantitative and a qualitative component, each with a separate rating. They will also run additional CloudResearch experiments to better understand how prior ratings and instructor gender impact learning outcomes and teaching ratings, across different time pressure conditions, and also how best to offer prior ratings information. Across these experiments, they will shed light on pitfalls in online teaching delivery that reduce equitable access, with a view to offering insight into how to improve online teaching delivery processes to improve equity, which is necessary in order to enable the broad reach of transformational new technologies.

Promising technologies like blockchain and AI can exacerbate disparities if not made accessible to all. Online courses on these subjects are vital for equitable access, but female instructors are underrepresented. On platforms like Udemy, only 23% of courses are taught by women, and in tech-intensive fields, this drops to 15%. Despite an increasing number of women enrolling in online courses, especially post-COVID-19 (e.g., 45% of total enrolments and 37% in STEM on Coursera in 2021), there is a need for inclusive teaching and learning environments. This project aims to address bias in ratings and improve equity through a collaboration with Coursera, using field experiments and additional online studies.